Structuring ordered nominal data for event sequence discovery.

Towards chip%2Don%2Dchip neuroscience fast mining of neuronal spike streams using graphics hardware.
Computational neuroscience is being revolutionized with the advent of multi-electrode arrays that provide real-time, dynamic perspectives into brain function. Mining neuronal spike streams from these chips is critical to understand the firing patterns of neurons and gain insight into the underlying cellular activity. To address this need, we present a solution that uses a massively parallel graphics processing unit (GPU) to mine the stream of spikes. We focus on mining frequent episodes that capture coordinated events across time even in the presence of intervening background events. Our contributions include new computation-to-core mapping schemes and novel strategies to map finite state machine-based counting algorithms onto the GPU. Together, these contributions move us towards a real-time ‘chip-onchip’ solution to neuroscience data mining, where one chip (the multi-electrode array) supplies the spike train data and another chip (the GPU) mines it at a scale previously unachievable.
Data Mining for Modeling Chiller Systems in Data Centers.

Diagnosing memory leaks using graph mining on heap dumps.
Memory leaks are caused by software programs that prevent the reclamation of memory that is no longer in use. They can cause significant slowdowns, exhaustion of available storage space and, eventually, application crashes. Detecting memory leaks is challenging because real-world applications are built on multiple layers of software frameworks, making it difficult for a developer to know whether observed references to objects are legitimate or the cause of a leak. We present a graph mining solution to this problem wherein we analyze heap dumps to automatically identify subgraphs which could represent potential memory leak sources. Although heap dumps are commonly analyzed in existing heap profiling tools, our work is the first to apply a graph grammar mining solution to this problem. Unlike classical graph mining work, we show that it suffices to mine the dominator tree of the heap dump, which is significantly smaller than the underlying graph. Our approach identifies not just leaking candidates and their structure, but also provides aggregate information about the access path to the leaks. We demonstrate several synthetic as well as real-world examples of heap dumps for which our approach provides more insight into the problem than state-of-the-art tools such as Eclipse’s MAT.
Extracting temporal signatures for comprehending systems biology models.
Systems biology has made massive strides in recent years, with capabilities to model complex systems including cell division, stress response, energy metabolism, and signaling pathways. Concomitant with their improved modeling capabilities, however, such biochemical network models have also become notoriously complex for humans to comprehend. We propose network comprehension as a key problem for the KDD community, where the goal is to create explainable representations of complex biological networks. We formulate this problem as one of extracting temporal signatures from multi-variate time series data, where the signatures are composed of ordinal comparisons between time series components. We show how such signatures can be inferred by formulating the data mining problem as one of feature selection in rank-order space. We propose five new feaure selection strategies for rank-order space and assess their selective superiorities. Experimental results on budding yeast cell cycle models demonstrate compelling results comparable to human interpretations of the cell cycle. Categories and Subject Descriptors G.3 [Probability and Statistics]: Time series analysis;
Unifying dependent clustering and disparate clustering for non%2Dhomogeneous data.
Modern data mining settings involve a combination of attributevalued descriptors over entities as well as specified relationships between these entities. We present an approach to cluster such non-homogeneous datasets by using the relationships to impose either dependent clustering or disparate clustering constraints. Unlike prior work that views constraints as boolean criteria, we present a formulation that allows constraints to be satisfied or violated in a smooth manner. This enables us to achieve dependent clustering and disparate clustering using the same optimization framework by merely maximizing versus minimizing the objective function. We present results on both synthetic data as well as several real-world datasets.
Efficiently Discovering Hammock Paths from Induced Similarity Networks

Program transformations for information personalization.
Personalization constitutes the mechanisms necessary to automatically customize information content, structure, and presentation to the end-user to reduce information overload. Unlike traditional approaches to personalization, the central theme of our approach is to model a website as a program and conduct website transformation for personalization by program transformation (e.g., partial evaluation, program slicing). The goal of this paper is study personalization through a program transformation lens, and develop a formal model, based on program transformations, for personalized interaction with hierarchical hypermedia. The specific research issues addressed involve identifying and developing program representations and transformations suitable for classes of hierarchical hypermedia, and providing supplemental interactions for improving the personalized experience. The primary form of personalization discussed is out-of-turn interaction – a technique which empowers a user navigating a hierarchical website to postpone clicking on any of the hyperlinks presented on the current page and, instead, communicate the
Dynamic tuning of algorithmic parameters of parallel scientific codes.

Mining linguistic cues for query expansion applications to drug interaction search.
Given a drug under development, what are other drugs or biochemical compounds that it might interact with? Early answers to this question, by mining the literature, are valuable for pharmaceutical companies, both monetarily and in avoiding public relations nightmares. Inferring drug-drug interactions is also important in designing combination therapies for complex diseases including cancers. We study this problem as one of mining linguistic cues for query expansion. By using (only) positive instances of drug interactions, we show how we can extract linguistic cues which can then be used to expand and reformulate queries to improve the effectiveness of drug interaction search. Our approach integrates many learning paradigms: partially supervised classification, association measures for collocation mining, and feature selection in supervised learning. We demonstrate compelling results on using positive examples from the DrugBank database to seed MEDLINE searches for drug interactions. In particular, we show that purely data-driven linguistic cues can be effectively mined and applied to realize a successful domain-specific query expansion framework.
Modular%2C Fine%2DGrained Adaptation of Parallel Programs.
Abstract. We present a modular approach to realizing fine-grained adaptation of program behavior in a parallel environment. Using a compositional framework based on function call interception and manipulation, the adaptive logic to monitor internal program states and control the behavior of program modules is written and managed as a separate code, thus supporting centralized design of complex adaptation strategies for adapting to dynamic changes within an application. By ‘catching ’ the functions that execute in synchronization across the parallel environment and inserting the adaptive logic operations at the intercepted control points, the proposed method provides a convenient way of synchronous adaptation without disturbing the parallel execution and communication structure already established in the original program. Applying our method to a CFD (computational fluid dynamics) simulation program to implement example adaptation scenarios, we demonstrate how effectively applications can change their behavior through fine-grained control. 1
Discovering Excitatory Networks from Discrete Event Streams with Applications to Neuronal Spike Train Analysis.
Abstract—Mining temporal network models from discrete event streams is an important problem with applications in computational neuroscience, physical plant diagnostics, and human-computer interaction modeling. We focus in this paper on temporal models representable as excitatory networks where all connections are stimulative, rather than inhibitive. Through this emphasis on excitatory networks, we show how they can be learned by creating bridges to frequent episode mining. Specifically, we show that frequent episodes help identify nodes with high mutual information relationships and which can be summarized into a dynamic Bayesian network (DBN). To demonstrate the practical feasibility of our approach, we show how excitatory networks can be inferred from both mathematical models of spiking neurons as well as real neuroscience datasets.


An Iterative Framework for the Improvement of Soft Clusters for Classificaction.

Sustainable operation and management of data center chillers using temporal data mining.
Motivation: Data centers are a critical component of modern IT infrastructure but are also among the worst environmental offenders through their increasing energy usage and the resulting large carbon footprints. Efficient management of data centers, including power management, networking, and cooling infrastructure, is hence crucial to sustainability. In the absence of a “first-principles ” approach to manage these complex components and their interactions, datadriven approaches have become attractive and tenable. Results: We present a temporal data mining solution to model and optimize performance of data center chillers, a key component of the cooling infrastructure. It helps bridge raw, numeric, time-series information from sensor streams toward higher level characterizations of chiller behavior, suitable for a data center engineer. To aid in this transduction, temporal data streams are first encoded into a symbolic representation, next run-length encoded segments are mined to form frequent motifs in time series, and finally these metrics are evaluated by their contributions to sustainability. A key innovation in our application is the ability to intersperse “don’t care ” transitions (e.g., transients) in continuous-valued time series data, an advantage we inherit by the application of frequent episode mining to symbolized representations of numeric time series. Our approach provides both qualitative and quantitative characterizations of the sensor streams to the data center engineer, to aid him in tuning chiller operating characteristics. This system is currently being prototyped for a data center managed by HP and experimental results from this application reveal the promise of our approach.
Accelerator%2DOriented Algorithm Transformation for Temporal Data Mining.
Abstract—Temporal data mining algorithms are becoming increasingly important in many application domains including computational neuroscience, especially the analysis of spike train data. While application scientists have been able to readily gather multi-neuronal datasets, analysis capabilities have lagged behind, due to both lack of powerful algorithms and inaccessibility to powerful hardware platforms. The advent of GPU architectures such as Nvidia’s GTX 280 offers a costeffective option to bring these capabilities to the neuroscientist’s desktop. Rather than port existing algorithms onto this architecture, we advocate the need for algorithm transformation, i.e., rethinking the design of the algorithm in a way that need not necessarily mirror its serial implementation strictly. We present a novel implementation of a frequent episode discovery algorithm by revisiting “in-the-large ” issues such as problem decomposition as well as “in-the-small ” issues such as data layouts and memory access patterns. This is non-trivial because frequent episode discovery does not lend itself to GPU-friendly data-parallel mapping strategies. Applications to many datasets and comparisons to CPU as well as prior GPU implementations showcase the advantages of our approach.
Recommender systems for the conference paper assignment problem.
We present a recommender systems approach to conference paper assignment, i.e., the task of assigning paper submissions to reviewers. We address both the modeling of reviewerpaper preferences (which can be cast as a learning problem) and the optimization of reviewing assignments to satisfy global conference criteria (which can be viewed as constraint satisfaction). Due to the paucity of preference data per reviewer or per paper (relative to other recommender systems applications) we show how we can integrate multiple sources of information to learn reviewer-paper preference models. Our models are evaluated not just in terms of prediction accuracy but in terms of end-assignment quality. Using a linear programming-based assignment optimization, we show how our approach better explores the space of unsupplied assignments to maximize the overall affinities of papers assigned to reviewers. We demonstrate our results on real reviewer bidding data from the IEEE ICDM 2007 conference.
Capturing truthiness mining truth tables in binary datasets.
We introduce a new data mining problem: mining truth tables in binary datasets. Given a matrix of objects and the properties they satisfy, a truth table identifies a subset of properties that exhibit maximal variability (and hence, complete independence) in occurrence patterns over the underlying objects. This problem is relevant in many domains, e.g., in bioinformatics where we seek to identify and model independent components of combinatorial regulatory pathways, and in social/economic demographics where we desire to determine independent behavioral attributes of populations. We outline a family of levelwise approaches adapted to mining truth tables, algorithmic optimizations, and applications to bioinformatics and political datasets.
Modular implementation of adaptive decisions in stochastic simulations.
We present a modular approach to implement adaptive decisions with existing scientific codes. Using a sophisticated system software tool based on the function call interception technique, an external code module is transparently combined with the given program without altering the original code structure, resulting in a newly composed application with extended behavior. This is useful for generalizing codes into using different parameter values or to switch algorithms or implementations at runtime. Applying the proposed method on a biochemical stochastic simulation software package to implement a set of exemplary use cases, which includes changing program parameters, substituting random number generators, and dynamically changing the stochastic simulation method, we demonstrate how effectively new code modules can be plugged in to construct an application with enhanced capabilities.


Towards Chip%2Don%2DChip Neuroscience Fast Mining of Frequent Episodes Using Graphics Processors

Accelerator%2DOriented Algorithm Transformation for Temporal Data Mining
Abstract—Temporal data mining algorithms are becoming increasingly important in many application domains including computational neuroscience, especially the analysis of spike train data. While application scientists have been able to readily gather multi-neuronal datasets, analysis capabilities have lagged behind, due to both lack of powerful algorithms and inaccessibility to powerful hardware platforms. The advent of GPU architectures such as Nvidia’s GTX 280 offers a costeffective option to bring these capabilities to the neuroscientist’s desktop. Rather than port existing algorithms onto this architecture, we advocate the need for algorithm transformation, i.e., rethinking the design of the algorithm in a way that need not necessarily mirror its serial implementation strictly. We present a novel implementation of a frequent episode discovery algorithm by revisiting “in-the-large ” issues such as problem decomposition as well as “in-the-small ” issues such as data layouts and memory access patterns. This is non-trivial because frequent episode discovery does not lend itself to GPU-friendly data-parallel mapping strategies. Applications to many datasets and comparisons to CPU as well as prior GPU implementations showcase the advantages of our approach.
Recommender Systems for the Conference Paper Assignment Problem
We present a recommender systems approach to conference paper assignment, i.e., the task of assigning paper submissions to reviewers. We address both the modeling of reviewerpaper preferences (which can be cast as a learning problem) and the optimization of reviewing assignments to satisfy global conference criteria (which can be viewed as constraint satisfaction). Due to the paucity of preference data per reviewer or per paper (relative to other recommender systems applications) we show how we can integrate multiple sources of information to learn reviewer-paper preference models. Our models are evaluated not just in terms of prediction accuracy but in terms of end-assignment quality. Using a linear programming-based assignment optimization, we show how our approach better explores the space of unsupplied assignments to maximize the overall affinities of papers assigned to reviewers. We demonstrate our results on real reviewer bidding data from the IEEE ICDM 2007 conference.
Computing with Proteins.
Protein Folding is considered one of today’s most significant “grand challenge ” problems and, together with other computational chemistry problems, has been one of the driving forces for the development of bigger, better, faster supercomputers. IBM is currently proposing to build the “Blue Gene ” [5], a petaflop computer to tackle the protein folding
The Pervasiveness of Data Mining asnd Machine Learning.

Temporal Process Discovery in Many Guises.

Protein Design by Sampling an Undirected Graphical Model of Residue Constraints.
Protein engineering seeks to produce amino acid sequences with desired characteristics, such as specified structure [1] or function [4]. This is a difficult problem due to interactions among residues; choosing an amino acid type at one position may constrain the possibilities at others, in order for the resulting protein to have proper structure and activity. To account for the dependence of some residues and take advantage of the independence of others, we have developed a new approach to protein design based on undirected probabilistic graphical models (Fig. 1). Our approach first constructs a graphical model that encodes residue constraints, and then uses the model generatively to produce new sequences optimized to meet the constraints. We focus here on constraints due to residue coupling, common pairs of amino acid types at particular pairs of positions, also known as correlated mutations or co-evolving residues. Recently, Ranganathan and colleagues showed that accounting for residue coupling, in addition to conservation, was to some extent both necessary and sufficient for viability of new WW domains [6, 5]. We have previously developed an approach for learning an undirected graphical model encapsulating conservation and coupling constraints in a protein family [7]. Our model provides a formal probabilistic semantics for reasoning about amino acid choices, defining a probability distribution function measuring how well a new sequence satisfies coupling constraints observed in the extant sequences of a family. Thus in order to design
Simultaneously Segmenting Multiple Gene Expression Time Courses by Analyzing Cluster Dynamics.
We present a new approach to segmenting multiple time series by analyzing the dynamics of cluster formation and re-arrangement around putative segment boundaries. This approach finds application in distilling large numbers of gene expression profiles into temporal relationships underlying biological processes. By directly minimizing information-theoretic measures of segmentation quality derived from Kullback-Leibler (KL) divergences, our formulation reveals clusters of genes along with a segmentation such that clusters show concerted behavior within segments but exhibit significant re-grouping across segmentation boundaries. The results of the segmentation algorithm can be summarized as Gantt charts revealing temporal dependencies in the ordering of key biological processes. Applications to the yeast metabolic cycle and the yeast cell cycle are described. 1.
Exploring Out%2Dof%2Dturn Interactions with Websites.

Simultaneously Segmenting Multiple Gene Expression Time Courses by Analyzing Cluster Dynamics.
We present a new approach to segmenting multiple time series by analyzing the dynamics of cluster formation and re-arrangement around putative segment boundaries. This approach finds application in distilling large numbers of gene expression profiles into temporal relationships underlying biological processes. By directly minimizing information-theoretic measures of segmentation quality derived from Kullback-Leibler (KL) divergences, our formulation reveals clusters of genes along with a segmentation such that clusters show concerted behavior within segments but exhibit significant re-grouping across segmentation boundaries. The results of the segmentation algorithm can be summarized as Gantt charts revealing temporal dependencies in the ordering of key biological processes. Applications to the yeast metabolic cycle and the yeast cell cycle are described. 1.
Seeded Discovery of Base Relations in Large Corpora.
Relationship discovery is the task of identifying salient relationships between named entities in text. We propose novel approaches for two sub-tasks of the problem: identifying the entities of interest, and partitioning and describing the relations based on their semantics. In particular, we show that term frequency patterns can be used effectively instead of supervised NER, and that the p-median clustering objective function naturally uncovers relation exemplars appropriate for describing the partitioning. Furthermore, we introduce a novel application of relationship discovery: the unsupervised identification of protein-protein interaction phrases. 1
Using Cost%2DSensitive Learning to Determine Gene Conversions.

Reconstructing chemical reaction networks data mining meets system identification.
We present an approach to reconstructing chemical reaction networks from time series measurements of the concentrations of the molecules involved. Our solution strategy combines techniques from numerical sensitivity analysis and probabilistic graphical models. By modeling a chemical reaction system as a Markov network (undirected graphical model), we show how systematically probing for sensitivities between molecular species can identify the topology of the network. Given the topology, our approach next uses detailed sensitivity profiles to characterize properties of reactions such as reversibility, enzyme-catalysis, and the precise stoichiometries of the reactants and products. We demonstrate applications to reconstructing key biological systems including the yeast cell cycle. In addition to network reconstruction, our algorithm finds applications in model reduction and model comprehension. We argue that our reconstruction algorithm can serve as an important primitive for data mining in systems biology applications.
Gaussian Process Models in Spatial Data Mining.

Unification of problem solving environment implementation layers with XML%2Dbased specifications.

Algorithms for Storytelling.
We formulate a new data mining problem called storytelling as a generalization of redescription mining. In traditional redescription mining, we are given a set of objects and a collection of subsets defined over these objects. The goal is to view the set system as a vocabulary and identify two expressions in this vocabulary that induce the same set of objects. Storytelling, on the other hand, aims to explicitly relate object sets that are disjoint (and hence, maximally dissimilar) by finding a chain of (approximate) redescriptions between the sets. This problem finds applications in bioinformatics, for instance, where the biologist is trying to relate a set of genes expressed in one experiment to another set, implicated in a different pathway. We outline an efficient storytelling implementation that embeds the CARTwheels redescription mining algorithm in an A * search procedure, using the former to supply next move operators on search branches to the latter. This approach is practical and effective for mining large datasets and, at the same time, exploits the structure of partitions imposed by the given vocabulary. Three application case studies are presented: a study of word overlaps in large English dictionaries, exploring connections between genesets in a bioinformatics dataset, and relating publications in the PubMed index of abstracts.
Graphical Models of Residue Coupling in Protein Families.
Identifying residue coupling relationships within a protein family can provide important insights into the family’s evolutionary record, and has significant applications in analyzing and optimizing sequence-structure-function relationships. We present the first algorithm to infer an undirected graphical model representing residue coupling in protein families. Such a model, which we call a residue coupling network, serves as a compact description of the joint amino acid distribution, focused on the independences among residues. This stands in contrast to current methods, which manipulate dense representations of co-variation and are focused on assessing dependence, which can conflate direct and indirect relationships. Our probabilistic model provides a sound basis for predictive (will this newly designed protein be folded and functional?), diagnostic (why is this protein not stable or functional?), and abductive reasoning (what if I attempt to graft features of one protein family onto another?). Further, our algorithm can readily incorporate, as priors, hypotheses regarding possible underlying mechanistic/energetic explanations for coupling. The resulting approach constitutes a powerful and discriminatory mechanism to identify residue coupling from protein sequences and structures. Analysis results on the G-protein coupled receptor (GPCR) and PDZ domain families demonstrate the ability of our approach to effectively uncover and exploit models of residue coupling.
Compositional specification and realisation of mixed%2Dinitiative web dialogs.

CMGSDB integrating heterogeneous Caenorhabditis elegans data sources using compositional data mining.
elegans data sources using compositional data mining
Biological data mining.
This paper presents an example of biological data mining: the recognition of promoters in DNA. We propose a two-level ensemble of classifiers to recognize E. Coli promoter sequences. The first-level classifiers include three Bayesian neural networks that learn from three different feature sets. The outputs of the first-level classifiers are combined in the second-level to give the final result. Empirical study shows that a precision rate of 92.2% is achieved, indicating an excellent performance of the proposed approach.
Compositional mining of multirelational biological datasets.

The Adaptive Code Kitchen Flexible Tools for Dynamic Application Composition.

From the Area Editor Search%2DThe New Incarnations.

Mining Web functional dependencies for flexible information access.
We present an approach to enhancing information access through web structure mining in contrast to traditional approaches involving usage mining. Specifically, we mine the hardwired hierarchical hyperlink structure of websites to identify patterns of term-term cooccurrences we call web FDs (functional dependencies). Intuitively, a web FD ‘x → y ’ declares that all paths through a site involving a hyperlink labeled x also contain a hyperlink labeled y. The complete set of FDs satisfied by a site help characterize (flexible and expressive) interaction paradigms supported by a site, where a paradigm is the set of explorable sequences therein. We describe algorithms for mining FDs, results from mining several hierarchical websites, and present several interface designs that can exploit such FDs to provide compelling user experiences. † Contact author 1
Validation and Estimation of Parameters for a General Probabilistic Model of the PCR Process.

S4W a problem%2Dsolving environment for wireless system design.

Scouts
This paper makes a case for paths as an explicit abstraction in operating system design. Paths provide a unifying infrastructure for several OS mechanisms that have been introduced in the last several years, including fbufs, integrated layer processing, packet classifiers, code specialization, and migrating threads. This paper articulates the potential advantages of a path-based OS structure, describes the specific path architecture implemented in the Scout OS, and demonstrates the advantages in a particular application domain---receiving, decoding, and displaying MPEG-compressed video. 1 Introduction  Layering is a fundamental structuring technique with a long history in system design. From early work on layered operating systems and network architectures [12, 32], to more recent advances in stackable systems [27, 15, 14, 26], layering has played a central role in managing complexity, isolating failure, and enhancing configurability. This paper describes a complementary, but equally f...
Scouts
This paper makes a case for paths as an explicit abstraction in operating system design. Paths provide a unifying infrastructure for several OS mechanisms that have been introduced in the last several years, including fbufs, integrated layer processing, packet classifiers, code specialization, and migrating threads. This paper articulates the potential advantages of a path-based OS structure, describes the specific path architecture implemented in the Scout OS, and demonstrates the advantages in a particular application domain---receiving, decoding, and displaying MPEG-compressed video. 1 Introduction  Layering is a fundamental structuring technique with a long history in system design. From early work on layered operating systems and network architectures [12, 32], to more recent advances in stackable systems [27, 15, 14, 26], layering has played a central role in managing complexity, isolating failure, and enhancing configurability. This paper describes a complementary, but equally f...
Scalable Execution of Legacy Scientific Codes.

Algorithms for storytelling.
We formulate a new data mining problem called storytelling as a generalization of redescription mining. In traditional redescription mining, we are given a set of objects and a collection of subsets defined over these objects. The goal is to view the set system as a vocabulary and identify two expressions in this vocabulary that induce the same set of objects. Storytelling, on the other hand, aims to explicitly relate object sets that are disjoint (and hence, maximally dissimilar) by finding a chain of (approximate) redescriptions between the sets. This problem finds applications in bioinformatics, for instance, where the biologist is trying to relate a set of genes expressed in one experiment to another set, implicated in a different pathway. We outline an efficient storytelling implementation that embeds the CARTwheels redescription mining algorithm in an A * search procedure, using the former to supply next move operators on search branches to the latter. This approach is practical and effective for mining large datasets and, at the same time, exploits the structure of partitions imposed by the given vocabulary. Three application case studies are presented: a study of word overlaps in large English dictionaries, exploring connections between genesets in a bioinformatics dataset, and relating publications in the PubMed index of abstracts.
BLOSOM a framework for mining arbitrary boolean expressions.
We introduce a novel framework (BLOSOM) for mining (frequent) boolean expressions  over binary-valued datasets. We organize the space of boolean expressions into  four categories: pure conjunctions, pure disjunctions, conjunction of disjunctions, and  disjunction of conjunctions. For each category, we propose a closure operator that  naturally leads to the concept of a closed boolean expression. The closed expressions  and their minimal generators give the most specific and most general boolean expressions  that are satisfied by their corresponding object set. Further, the closed/minimal  generator expressions form a lossless representation of all possible boolean expressions. BLOSSOM efficiently
Nonorthogonal decomposition of binary matrices for bounded%2Derror data compression and analysis.
This article presents the design and implementation of a software tool, PROXIMUS, for error-bounded approximation of high-dimensional binary attributed datasets based on nonorthogonal decomposition of binary matrices. This tool can be used for analyzing data arising in a variety of domains ranging from commercial to scientific applications. Using a combination of innovative algorithms, novel data structures, and efficient implementation, PROXIMUS demonstrates excellent accuracy, performance, and scalability to large datasets. We experimentally demonstrate these on diverse applications in association rule mining and DNA microarray analysis. In limited beta release, PROXIMUS currently has over 300 installations in over 10 countries.
A general probabilistic model of the PCR process.

The JigCell Model Builder A Spreadsheet Interface for Creating Biochemical Reaction Network Models.
Abstract—Converting a biochemical reaction network to a set of kinetic rate equations is tedious and error prone. We describe known interface paradigms for inputing models of intracellular regulatory networks: graphical layout (diagrams), wizards, scripting languages, and direct entry of chemical equations. We present the JigCell Model Builder, which allows users to define models as a set of reaction equations using a spreadsheet (an example of direct entry of equations) and outputs model definitions in the Systems Biology Markup Language, Level 2. We present the results of two usability studies. The spreadsheet paradigm demonstrated its effectiveness in reducing the number of errors made by modelers when compared to hand conversion of a wiring diagram to differential equations. A comparison of representatives of the four interface paradigms for a simple model of the cell cycle was conducted which measured time, mouse clicks, and keystrokes to enter the model, and the number of screens needed to view the contents of the model. All four paradigms had similar data entry times. The spreadsheet and scripting language approaches require significantly fewer screens to view the models than do the wizard or graphical layout approaches. Index Terms—Biochemical reaction networks, bioinformatics, modeling, user interface paradigms. 1
Interacting with Web Hierarchies.
This paper proposes a new visualization and interaction technique for medium-sized trees, called Collapsible Cylindrical Trees (CCT). Child nodes are mapped on rotating cylinders, which will be dynamically displayed or hidden to achieve a useful balance of detail and context. Besides a comprehensible three-dimensional visualization of trees, the main feature of CCT is a very fast and intuitive interaction with the displayed nodes. Just one click is needed to reach every node and perform an action on it such as displaying a web page. The CCT browsing technique was developed for interaction with web hierarchies but is not limited to this domain. We also present sample implementations of CCT using VRML, which show the usefulness of this intuitive tree navigation technique.
Spatial data mining to support pandemic preparedness.

Network reconstruction from dynamic data.

Redescription Mining Structure Theory and Algorithms.
We introduce a new data mining problem—redescription mining—that unifies considerations of conceptual clustering, constructive induction, and logical formula discovery. Redescription mining begins with a collection of sets, views it as a propositional vocabulary, and identifies clusters of data that can be defined in at least two ways using this vocabulary. The primary contributions of this paper are conceptual and theoretical: (i) we formally study the space of redescriptions underlying a dataset and characterize their intrinsic structure, (ii) we identify impossibility as well as strong possibility results about when mining redescriptions is feasible, (iii) we present several scenarios of how we can custom-build redescription mining solutions for various biases, and (iv) we outline how many problems studied in the larger machine learning community are really special cases of redescription mining. By highlighting its broad scope and relevance, we aim to establish the importance of redescription mining and make the case for a thrust in this new line of research.
A Generative Programming Approach to Interactive Information Retrieval Insights and Experiences.
Abstract. We describe the application of generative programming to a problem in interactive information retrieval. The particular interactive information retrieval problem we study is the support for ‘out of turn interaction ’ with a website – how a user can communicate input to a website when the site is not soliciting such information on the current page, but will do so on a subsequent page. Our solution approach makes generous use of program transformations (partial evaluation, currying, and slicing) to delay the site’s current solicitation for input until after the user’s out-of-turn input is processed. We illustrate how studying outof-turn interaction through a generative lens leads to several valuable insights: (i) the concept of a web dialog, (ii) an improved understanding of web taxonomies, and (iii) new web interaction techniques and interfaces. These notions allow us to cast the design of interactive (and responsive) websites in terms of the underlying dialog structure and, further, suggest a simple implementation strategy with a clean separation of concerns. We also highlight new research directions opened up by the generative programming approach to interactive information retrieval such as the idea of web interaction axioms. 1
Reasoning about sets using redescription mining.
Redescription mining is a newly introduced data mining problem that seeks to find subsets of data that afford multiple definitions. It can be viewed as a generalization of association rule mining, from finding implications to equivalences; as a form of conceptual clustering, where the goal is to identify clusters that afford dual characterizations; and as a form of constructive induction, to build features based on given descriptors that mutually reinforce each other. In this paper, we present the use of redescription mining as an important tool to reason about a collection of sets, especially their overlaps, similarities, and differences. We outline algorithms to mine all minimal (non-redundant) redescriptions underlying a dataset using notions of minimal generators of closed itemsets. We also show the use of these algorithms in an interactive context, supporting constraint-based exploration and querying. Specifically, we showcase a bioinformatics application that empowers the biologist to define a vocabulary of sets underlying a domain of genes and to reason about these sets, yielding significant biological insight.
Gaussian Processes for Active Data Mining of Spatial Aggregates.
We present an active data mining mechanism for qualitative  analysis of spatial datasets, integrating identification  and analysis of structures in spatial data with targeted  collection of additional samples. The mechanism  is designed around the spatial aggregation language  (SAL) for qualitative spatial reasoning, and seeks to uncover  high-level spatial structures from only a sparse set  of samples. This approach is important for applications  in domains such as aircraft design, wireless system simulation,  fluid dynamics, and sensor networks. The mechanism  employs Gaussian processes, a formal mathematical  model for reasoning about spatial data, in order to  build surrogate models from sparse data, reason about  the uncertainty of estimation at unsampled points, and  formulate objective criteria for closing-the-loop between  data collection and data analysis. It optimizes sample  selection using entropy-based functionals defined over  spatial aggregates instead of the traditional approach  of sampling to minimize estimated variance. We apply  this mechanism on a global optimization benchmark  comprising a testbank of 2D functions, as well as on  data from wireless system simulations. The results reveal  that the proposed sampling strategy makes more  judicious use of data points by selecting locations that  clarify high-level structures in data, rather than choosing  points that merely improve quality of function approximation.
A Model Based Approach to Reachability Routing

Novel runtime systems support for adaptive compositional modeling in PSEs.
Problem solving environments (PSEs) have progressed significantly in the past few years. The vision of truly seamless PSEs relies on runtime systems support that is cognizant of the operational issues underlying scientific computations and, at the same time, is flexible enough to accommodate diverse application scenarios. This paper presents a PSE runtime support solution through a novel combination of two computational technologies – Weaves, a source-language independent parallel runtime compositional framework that operates through reverse-analysis of compiled object files, and runtime recommender systems that aid in dynamic knowledge-based application composition. Domain-specific adaptivity is exploited through runtime recommendation of code modules and a sophisticated checkpointing framework for transparent deployment. A core set of “adaptivity schemas ” are provided as templates for adaptive composition of large-scale scientific computations. Implementation issues, motivating application contexts, and preliminary results are described. 1
From The Area Editor Frontiers of Search.

Compression
We study the notion of Compressed Sensing (CS) as put forward in [14] and related work [20, 3, 4]. The basic idea behind CS is that a signal or image, unknown but supposed to be compressible by a known transform, (eg. wavelet or Fourier), can be subjected to fewer measurements than the nominal number of pixels, and yet be accurately reconstructed. The samples are nonadaptive and measure ‘random ’ linear combinations of the transform coefficients. Approximate reconstruction is obtained by solving for the transform coefficients consistent with measured data and having the smallest possible ℓ 1 norm. We perform a series of numerical experiments which validate in general terms the basic idea proposed in [14, 3, 5], in the favorable case where the transform coefficients are sparse in the strong sense that the vast majority are zero. We then consider a range of less-favorable cases, in which the object has all coefficients nonzero, but the coefficients obey an ℓ p bound, for some p ∈ (0, 1]. These experiments show that the basic inequalities behind the CS method seem to involve reasonable constants. We next consider synthetic examples modelling problems in spectroscopy and image processing, and note that reconstructions from CS are often visually “noisy ”. We post-process using translation-invariant de-noising, and find the visual appearance considerably improved. We also consider a multiscale deployment of compressed sensing, in which various scales are segregated and CS applied separately to each; this gives much better quality reconstructions than a literal deployment of the CS methodology. We also report that several workable families of ‘random ’ linear combinations all behave equivalently, including random spherical, random signs, partial Fourier and partial Hadamard. These results show that, when appropriately deployed in a favorable setting, the CS framework is able to save significantly over traditional sampling, and there are many useful extensions of the basic idea.
Spatial Aggregation for Qualitative Assessment of Scientific Computations.
Qualitative assessment of scientific computations is an emerging application area that applies a data-driven approach to characterize, at a high level, phenomena including conditioning of matrices, sensitivity to various types of error propagation, and algorithmic convergence behavior. This paper develops a spatial aggregation approach that formalizes such analysis in terms of model selection utilizing spatial structures extracted from matrix perturbation datasets. We focus in particular on the characterization of matrix eigenstructure, both analyzing sensitivity of computations with spectral portraits and determining eigenvalue multiplicity with Jordan portraits. Our approach employs spatial reasoning to overcome noise and sparsity by detecting mutually reinforcing interpretations, and to guide subsequent data sampling. It enables quantitative evaluation of properties of a scientific computation in terms of confidence in a model, explainable in terms of the sampled data and domain knowledge about the underlying mathematical structure. Not only is our methodology more rigorous than the common approach of visual inspection, but it also is often substantially more efficient, due to well-defined stopping criteria. Results show that the mechanism efficiently samples perturbation space and successfully uncovers high-level properties of matrices.
Staging transformations for multimodal web interaction management.
Multimodal interfaces are becoming increasingly ubiquitous with the advent of mobile devices, accessibility considerations, and novel software technologies that combine diverse interaction media. In addition to improving access and delivery capabilities, such interfaces enable flexible and personalized dialogs with websites, much like a conversation between humans. In this paper, we present a software framework for multimodal web interaction management that supports mixed-initiative dialogs between users and websites. A mixed-initiative dialog is one where the user and the website take turns changing the flow of interaction. The framework supports the functional specification and realization of such dialogs using staging transformations -- a theory for representing and reasoning about dialogs based on partial input. It supports multiple interaction interfaces, and offers sessioning, caching, and co-ordination functions through the use of an interaction manager. Two case studies are presented to illustrate the promise of this approach.
Enhancing usability in CITIDEL multimodal
We describe four usability-enhancing interfaces to CITIDEL aimed at improving the user experience and supporting personalized information access by targeted communities. These comprise: a multimodal interaction facility with capability for out-of-turn input, interactive visualizations for exploratory analysis, a translation center exposing multilingual interfaces, as well as traditional usability enhancements. Pilot studies demonstrate the resulting improvements in quality, as measured across a number of metrics. Categories and Subject Descriptors H.5.2 [Information Interfaces and Presentation]: User
Turning CARTwheels an alternating algorithm for mining redescriptions.

Staging transformations for multimodal web interaction management.
Multimodal interfaces are becoming increasingly ubiquitous with the advent of mobile devices, accessibility considerations, and novel software technologies that combine diverse interaction media. In addition to improving access and delivery capabilities, such interfaces enable flexible and personalized dialogs with websites, much like a conversation between humans. In this paper, we present a software framework for multimodal web interaction management that supports mixed-initiative dialogs between users and websites. A mixed-initiative dialog is one where the user and the website take turns changing the flow of interaction. The framework supports the functional specification and realization of such dialogs using staging transformations -- a theory for representing and reasoning about dialogs based on partial input. It supports multiple interaction interfaces, and offers sessioning, caching, and co-ordination functions through the use of an interaction manager. Two case studies are presented to illustrate the promise of this approach.
The JigCell Model Builder and Run Manager.
Summary: We describe the JigCell Model Builder (JCMB), a tool for creating biochemical reaction network models. JCMB is designed for ease of use and its interface uses the standard spreadsheet metaphor. The JigCell Run Manager (JCRM) is a tool for organizing the large collections of simulation runs typically required by reaction network modeling activities. Availability: JCMB and JCRM are part of the JigCell suite
Automatically Generating Interfaces for Personalized Interaction with Digital Libraries

Globally optimal transmitter placement for indoor wireless communication systems.
The DIRECT (DIviding RECTangles) algorithm [23], a variant of Lipschitzian methods for bound constrained global optimization, has been applied to the optimal transmitter placement for indoor wireless systems. Power coverage and BER (bit error rate) are considered as two criteria for optimizing locations of a specified number of transmitters across the feasible region of the design space. The performance of a DIRECT implementation in such applications depends on the characteristics of the objective function, the problem dimension, and the desired solution accuracy. Implementations with static data structures often fail in practice because of unpredictable memory requirements. This is especially critical in S 4 W (Site-Specific System Simulator for Wireless communication systems), where the DIRECT optimization is just one small component connected to a parallel 3D propagation ray tracing modeler running on a 200-node Beowulf cluster of Linux workstations, and surrogate functions for a WCDMA (wideband code division multiple access) simulator are also used to estimate the channel performance. Any component failure of this large computation would abort the entire design process. To make the DIRECT global optimization algorithm efficient and robust, a set of dynamic data structures is proposed here to balance the memory requirements with execution time, while simultaneously adapting to arbitrary problem size. The focus is on design issues of the dynamic data structures, related memory management strategies, and application issues of the DIRECT algorithm to the transmitter placement optimization for wireless communication systems. Results for two indoor systems are presented to demonstrate the effectiveness of the present work.
Personalization by Program Slicing.
Personalization involves customizing information access to the end-user. As any new area of computer science research it lacks formal models to guide the design of systems. In this paper, we present a modeling methodology, based on generative programming, for personalizing interactions with hierarchical websites. The methodology entails modeling a user’s interaction with a site in a program and applying program slicing to personalize the interaction. While preserving interactivity, this approach does not require the designer to anticipate all possible user interactions a priori and provide interfaces for each. Moreover, it provides a theoretical, systematic, and implementationneutral way to design systems and is therefore a timely contribution to the young field of personalization as well as a novel application of generative programming. 1
Towards integrated PSEs for wireless communications experiences with the S4W and SitePlanner%AE projects.

Gaussian Process Models of Spatial Aggregation Algorithms.
Multi-level spatial aggregates are important for data mining in a variety of scientific and engineering applications, from analysis of weather data (aggregating temperature and pressure data into ridges and fronts) to performance analysis of wireless systems (aggregating simulation results into configuration space regions exhibiting particular performance characteristics). In many of these applications, data collection is expensive and time consuming, so effort must be focused on gathering samples at locations that will be most important for the analysis. This requires that we be able to functionally model a data mining algorithm in order to assess the impact of potential samples on the mining of suitable spatial aggregates. This paper describes a novel Gaussian process approach to modeling multi-layer spatial aggregation algorithms, and demonstrates the ability of the resulting models to capture the essential underlying qualitative behaviors of the algorithms. By helping cast classical spatial aggregation algorithms in a rigorous quantitative framework, the Gaussian process models support diverse uses such as directed sampling, characterizing the sensitivity of a mining algorithm to particular parameters, and understanding how variations in input data fields percolate up through a spatial aggregation hierarchy. 1
Expresso and Chips Creating a Next Generation Microarray Experiment Management System.
Expresso is an experiment management system that is designed to assist biologists in planning, executing, and interpreting microarray experiments. It serves as a unifying framework to study data-driven application composition systems, as envisaged under the NSF Next Generation Software (NGS) program. Physical and analytical stages of the microarray process are mirrored in Expresso with computational models from biophysics, molecular biology, biochemistry, robotics, image processing, statistics, and knowledge representation. These models are pushed deeper (earlier) into the design process to help avoid costly design errors and to provide, as needed, surrogate functions for the traditional stages of microarray experiments. In this paper, we describe ongoing work in the design of Expresso, with specific reference to application composition, application optimization, experiment protocol design, and `closing the loop.&#039;  1. 
The Traits of the Personable.
Abstract. Information personalization is fertile ground for application of AI techniques. In this article I relate personalization to the ability to capture partial information in an information-seeking interaction. The specific focus is on personalizing interactions at web sites. Using ideas from partial evaluation and explanation-based generalization, I present a modeling methodology for reasoning about personalization. This approach helps identify seven tiers of ‘personable traits ’ in web sites. 1
Next generation modeling II %2D applications improving the development process for eukaryotic cell cycle models with a modeling support environment.

Personalizing interactions with information systems.
Personalization constitutes the mechanisms and technologies necessary to customize information access to the end-user. It can be defined as the automatic adjustment of information content, structure, and presentation tailored to the individual. In this chapter, we study personalization from the viewpoint of personalizing interaction. The survey covers mechanisms for information-finding on the web, advanced information retrieval systems, dialogbased applications, and mobile access paradigms. Specific emphasis is placed on studying how users interact with an information system and how the system can encourage and foster interaction. This helps bring out the role of the personalization system as a facilitator which reconciles the user&#039;s mental model with the underlying information system&#039;s organization. Three tiers of personalization systems are presented, paying careful attention to interaction considerations. These tiers show how progressive levels of sophistication in interaction can be achieved. The chapter also surveys systems support technologies and niche application domains.
Novel Runtime Systems Support for Adaptive Compositional Modeling on the Grid

Turning CARTwheels An Alternating Algorithm for Mining Redescriptions

Taking the Initiative with Extempore Exploring Out%2Dof%2DTurn Interactions with Websites

Supporting Out%2Dof%2Dturn Interactions in a Multimodal Web Interface

Staging Transformations for Multimodal Web Interaction Management
Multimodal interfaces are becoming increasingly ubiquitous with the advent of mobile devices, accessibility considerations, and novel software technologies that combine diverse interaction media. In addition to improving access and delivery capabilities, such interfaces enable flexible and personalized dialogs with websites, much like a conversation between humans. In this paper, we present a software framework for multimodal web interaction management that supports mixed-initiative dialogs between users and websites. A mixed-initiative dialog is one where the user and the website take turns changing the flow of interaction. The framework supports the functional specification and realization of such dialogs using staging transformations -- a theory for representing and reasoning about dialogs based on partial input. It supports multiple interaction interfaces, and offers sessioning, caching, and co-ordination functions through the use of an interaction manager. Two case studies are presented to illustrate the promise of this approach.
Reinforcing reachable route.
This paper studies the evaluation of routing algorithms from the perspective of reachability routing, where the  goal is to determine all paths between a sender and a receiver. Reachability routing is becoming relevant with the  changing dynamics of the Internet and the emergence of low-bandwidth wireless/ad-hoc networks. We make the  case for reinforcement learning as the framework of choice to realize reachability routing, within the confines of  the current Internet infrastructure. The setting of the reinforcement learning problem offers several advantages,  including loop resolution, multi-path forwarding capability, cost-sensitive routing, and minimizing state overhead,  while maintaining the incremental spirit of current backbone routing algorithms. We identify research issues  in reinforcement learning applied to the reachability routing problem to achieve a fluid and robust backbone  routing framework. This paper also presents the design, implementation and evaluation of a new reachability  routing algorithm that uses a model-based approach to achieve cost-sensitive multi-path forwarding; performance  assessment of the algorithm in various troublesome topologies shows consistently superior performance over  classical reinforcement learning algorithms. The paper is targeted toward practitioners seeking to implement a  reachability routing algorithm.
From landscapes to waterscapes A PSE for landuse change analysis.
We describe the design and implementation of L2W- a problem solving environment (PSE) for landuse change analysis. L2W organizes and unifies the diverse collection of software typically associated with ecosystem models (hydrological, economic, and biological). It provides a web-based interface for potential watershed managers and other users to explore meaningful alternative land development and management scenarios and view their hydrological, ecological, and economic impacts. A prototype implementation
Studying Recommendation Algorithms by Graph Analysis.
We present a novel framework for studying recommendation algorithms in terms of the  `jumps&#039; that they make to connect people to artifacts. This approach emphasizes reachability  via an algorithm within the implicit graph structure underlying a recommender  dataset and allows us to consider questions relating algorithmic parameters to properties  of the datasets. For instance, given a particular algorithm `jump,&#039; what is the average  path length from a person to an artifact? Or, what choices of minimum ratings and  jumps maintain a connected graph? We illustrate the approach with a common jump  called the `hammock&#039; using movie recommender datasets.
BSML A binding schema markup language for data interchange in problem solving environments.
We describe a binding schema markup language (BSML) for describing data interchange between scientific codes. Such a facility is an important constituent of scientific problem solving environments (PSEs). BSML is designed to integrate with a PSE or application composition system that views model specification and execution as a problem of managing semistructured data. The data interchange problem is addressed by three techniques for processing semistructured data: validation, binding, and conversion. We present BSML and describe its application to a PSE for wireless communications system design. 1
Improving the Development Process for Eukaryotic Cell Cycle Models with a Modeling Support Environment.
On behalf of:
S4W Globally Optimized Design of Wireless Communications Systems.
In this paper, a global optimization technique is applied to solve the optimal transmitter placement problem for indoor wireless systems. An e#cient pattern search algorithm---DIRECT (DIviding RECTangles) of Jones, Perttunen, and Stuckman (1993)---has been connected to a parallel 3D radio propagation ray tracing modeler running on a 200-node Beowulf cluster of Linux workstations. Surrogate functions for a parallel WCDMA (wideband code division multiple access) simulator were used to estimate the system performance for the global optimization algorithm. Power coverage and BER (bit error rate) are considered as two di#erent criteria for optimizing locations of a specified number of transmitters across the feasible region of the design space. This paper briefly describes the underlying radio propagation and WCDMA simulations and focuses on the design issues of the optimization loop.
Algebraic Techniques for Analysis of Large Discrete%2DValued Datasets.
With the availability of large scale computing platforms and instrumentation  for data gathering, increased emphasis is being placed on efficient  techniques for analyzing large and extremely high-dimensional datasets. In this  paper, we present a novel algebraic technique based on a variant of semi-discrete  matrix decomposition (SDD), which is capable of compressing large discretevalued  datasets in an error bounded fashion. We show that this process of compression  can be thought of as identifying dominant patterns in underlying data.
Qualitative Analysis of Correspondence for Experimental Algorithmics

The Traits of the Personable
Abstract. Information personalization is fertile ground for application of AI techniques. In this article I relate personalization to the ability to capture partial information in an information-seeking interaction. The specific focus is on personalizing interactions at web sites. Using ideas from partial evaluation and explanation-based generalization, I present a modeling methodology for reasoning about personalization. This approach helps identify seven tiers of ‘personable traits ’ in web sites. 1
BSML A Binding Schema Markup Language for Data Interchange in Problem Solving Environments %28PSEs%29

Sampling Strategies for Mining in Data%2DScarce Domains
description Redescribe Redescribe Update objects N-graph classes Equivalence Spatial Figure 2. The Spatial Aggregation Language&#039;s multilayer spatial aggregates, uncovered by a uniform vocabulary of operators using domain knowledge. We can express several scientific data mining tasks---such as vector field bundling, contour aggregation, correspondence abstraction, clustering, and uncovering regions of uniformity---as multilevel computations with SAL aggregates. of the hierarchy (streamlines bundled with convergent flows). The SALpLCPMK compMKM different streamline aggregations from a neighborhood grap and chooses one on the basis of how well its curvature matches the direction of the vectors it aggregates. If data is scarce, some of these classification decisions could be ambiguous ---multipP streamline aggregations might exist. In such a case, we would want to choose a new datasampM that reduces the ambiguity and clarifies what the correct classification should be. This is the essence of oursampLLE methodology: using SAL aggregates, we identify an information -theoretic measure (here, ambiguity) that can drive stages of future data collection. For instance, we can summarize the ambiguous streamline classifications as a 2D ambiguity distribution that has aspAL for every location where we detected an ambiguity. Ambiguity reduction is apMEELE of minimizing (or maximizing, as the case may be) a functional involving thecompNKN ambiguity. The functional could be the underlying data field&#039;sentrop , as the ambiguity distribution reveals. Such a minimization will lead us to select a data ptaC that clarifies the distribution of streamlines, and hence that more effectively uses data for data miningpningCLR This methodology&#039;s net effect is that we cancapLBM apPMBPCDPY design&#039;s desi...
Using Hierarchical Data Mining to Characterize Performance of Wireless System Configurations
This paper presents a statistical framework for assessing wireless systems performance using hierarchical data mining techniques. We consider WCDMA (wideband code division multiple access) systems with two-branch STTD (space time transmit diversity) and 1/2 rate convolutional coding (forward error correction codes). Monte Carlo simulation estimates the bit error probability (BEP) of the system across a wide range of signal-to-noise ratios (SNRs). A performance database of simulation runs is collected over a targeted space of system configurations. This database is then mined to obtain regions of the configuration space that exhibit acceptable average performance. The shape of the mined regions illustrates the joint influence of configuration parameters on system performance. The role of data mining in this application is to provide explainable and statistically valid design conclusions. The research issue is to define statistically meaningful aggregation of data in a manner that permits efficient and effective data mining algorithms. We achieve a good compromise between these goals and help establish the applicability of data mining for characterizing wireless systems performance. 1
Weaves A Novel Direct Code Execution Interface for Parallel High Performance Scientific Codes

Reinforcing Reachable Routes
This paper studies the evaluation of routing algorithms from the perspective of reachability routing, where the  goal is to determine all paths between a sender and a receiver. Reachability routing is becoming relevant with the  changing dynamics of the Internet and the emergence of low-bandwidth wireless/ad-hoc networks. We make the  case for reinforcement learning as the framework of choice to realize reachability routing, within the confines of  the current Internet infrastructure. The setting of the reinforcement learning problem offers several advantages,  including loop resolution, multi-path forwarding capability, cost-sensitive routing, and minimizing state overhead,  while maintaining the incremental spirit of current backbone routing algorithms. We identify research issues  in reinforcement learning applied to the reachability routing problem to achieve a fluid and robust backbone  routing framework. This paper also presents the design, implementation and evaluation of a new reachability  routing algorithm that uses a model-based approach to achieve cost-sensitive multi-path forwarding; performance  assessment of the algorithm in various troublesome topologies shows consistently superior performance over  classical reinforcement learning algorithms. The paper is targeted toward practitioners seeking to implement a  reachability routing algorithm.
Programming environments for multidisciplinary Grid communities.
this paper is on supporting multidisciplinary communities of scientists and engineers. We discuss requirements for Grid computing environments (GCEs) in this context, and describe several core support technologies developed to meet these requirements. Our work extends the notion of a programming environment beyond the compile--schedule--execute paradigm, to include functionality such as collaborative application composition, information services, and data and simulation management. Systems designed for five different applications communities are described. These systems illustrate common needs and characteristics arising in multidisciplinary communities and motivate a highlevel design framework for building GCEs that meet those needs. Copyright # 2002 John Wiley &amp; Sons, Ltd
MyPYTHIA a recommendation portal for scientific software and services.
In this paper, we outline the design of a recommendation system (MyPYTHIA) implemented as a web portal.  MyPYTHIA&#039;s design objectives include evaluating the quality and performance of scientific software on grid  platforms, creating knowledge about which software and computational services should be selected for solving  particular problems, selecting parameters of software (or of computational services) based on user-specified  computational objectives, providing access to performance data and knowledge bases over the web, and enabling  recommendations for targeted application domains. MyPYTHIA uses a combination of statistical analysis, pattern  extraction techniques, and a database of software performance to map feature-based representations of problem  instances to appropriate software. MyPYTHIA&#039;s open architecture allows the user to customize it for conducting  individual case studies. We describe the architecture as well as several scientific domains of knowledge enabled  by such case studies.    This work was supported in part by NSF grant CDA 91-23502, PRF 6902851, DARPA grant N66001-97-C-8533 (Navy), DOE LG-6982, DARPA under ARO grant DAAH04-94-G-0010, and the Purdue Research Foundation.  1  Contents  1 
The Emerging Landscape of Bioinformatics Software Systems.

Modeling and Building Personalized Digital Libraries with PIPE and 5SL.

Ambiguity%2DDirected Sampling for Qualitative Analysis of Sparse Data from Spatially%2DDistributed Physical Systems.
A number of important scientific and engineering
Compression of particle data from hierarchical approximate methods.
This article presents an analytical and computational framework for the compression of particle data resulting from hierarchical approximate treecodes such as the Barnes-Hut and Fast Multipole Methods. Due to approximations introduced by hierarchical methods, various parameters (such as position, velocity, acceleration, potential) associated with a particle can be bounded by distortion radii. Using this distortion radii, we develop storage schemes that guarantee error bounds while maximizing compression. Our schemes make extensive use of spatial and temporal coherence of particle behavior and yield compression ratios higher than 12:1 over raw data, and 6:1 over gzipped (LZ) raw data for selected simulation instances. We demonstrate that for uniform distributions with 2M particles, storage requirements can be reduced from 24 MB to about 1.8 MB (about 7 bits per particle per timestep) for storing particle positions. This is significant because it enables faster storage/retrieval, better temporal resolution, and improved analysis. Our results are shown to scale from small systems (2K particles) to much larger systems (over 2M particles). The associated...
Mining scientific data.
this article, we give an
When being Weak is Brave Privacy in Recommender Systems

Explaining Scenarios for Information Personalization
Personalization customizes information access. The PIPE (`Personalization is Partial Evaluation&#039;) modeling methodology represents interaction with an information space as a program. The program is then specialized to a user&#039;s known interests or information seeking activity by the technique of partial evaluation. In this paper, we elaborate PIPE by considering requirements analysis in the personalization lifecycle. We investigate the use of scenarios as a means of identifying and analyzing personalization requirements. As our first result, we show how designing a PIPE representation can be cast as a search within a space of PIPE models, organized along a partial order. This allows us to view the design of a personalization system, itself, as specialized interpretation of an information space. We then exploit the underlying equivalence of explanation-based generalization (EBG) and partial evaluation to realize high-level goals and needs identified in scenarios
Evaluating Recommendation Algorithms by Graph Analysis
We present a novel framework for evaluating recommendation algorithms in terms of the  `jumps&#039; that they make to connect people to artifacts. This approach emphasizes reachability  via an algorithm within the implicit graph structure underlying a recommender  dataset, and serves as a complement to evaluation in terms of predictive accuracy. The  framework allows us to consider questions relating algorithmic parameters to properties  of the datasets. For instance, given a particular algorithm `jump,&#039; what is the average  path length from a person to an artifact? Or, what choices of minimum ratings and  jumps maintain a connected graph? We illustrate the approach with a common jump  called the `hammock&#039; using movie recommender datasets.
The Partial Evaluation Approach to Information Personalization
Information personalization refers to the automatic adjustment of information content, structure, and presentation tailored to an individual user. By reducing information overload and customizing information access, personalization systems have emerged as an important segment of the Internet economy. This paper presents a systematic modeling methodology --- PIPE (`Personalization is Partial Evaluation&#039;) --- for personalization. Personalization systems are designed and implemented in PIPE by modeling an information-seeking interaction in a programmatic representation. The representation supports the description of information-seeking activities as partial information and their subsequent realization by partial evaluation, a technique for specializing programs. We describe the modeling methodology at a conceptual level and outline representational choices. We present two application case studies that use PIPE for personalizing web sites and describe how PIPE suggests a novel evaluation criterion for information system designs. Finally, we mention several fundamental implications of adopting the PIPE model for personalization and when it is (and is not) applicable.
The Expresso Microarray Experiment Management System The Functional Genomics of Stress Responses in Loblolly Pine
:  Conception, design, and implementation of cDNA microarray experiments present a variety of bioinformatics challenges for biologists and computational scientists. The multiple stages of data acquisition and analysis have motivated the design of Expresso, a system for microarray experiment management. Salient aspects of Expresso include support for clone replication and randomized placement; automatic gridding, extraction of expression data from each spot, and quality monitoring; flexible methods of combining data from individual spots into information about clones and functional categories; and the use of inductive logic programming for higher-level data analysis and mining. The development of Expresso is occurring in parallel with several generations of microarray experiments aimed at elucidating genomic responses to drought stress in loblolly pine seedlings. The current experimental design incorporates 384 pine cDNAs replicated and randomly placed in two specific microarray layouts. We describe the design of Expresso as well as results of analysis with Expresso that suggest the importance of molecular chaperones and membrane transport proteins in mechanisms conferring successful adaptation to long-term drought stress.  
Mixed%2DInitiative Interaction %3D Mixed Computation
We show that partial evaluation can be usefully viewed as a programming model for realizing mixed-initiative  functionality in interactive applications. Mixed-initiative interaction between two participants is one where the  parties can take turns at any time to change and steer the flow of interaction. We concentrate on the facet of  mixed-initiative referred to as `unsolicited reporting&#039; and demonstrate how out-of-turn interactions by users can  be modeled by `jumping ahead&#039; to nested dialogs (via partial evaluation). Our approach permits the view of  dialog management systems in terms of their native support for staging and simplifying interactions; we characterize  three different voice-based interaction technologies using this viewpoint. In particular, we show that the  built-in form interpretation algorithm (FIA) in the VoiceXML dialog management architecture is actually a (well  disguised) combination of an interpreter and a partial evaluator.    This work is supported in part by US National Science Foundation grants DGE-9553458 and IIS-9876167.  1  1 
Privacy Risks in Recommender Systems.
this article, we can state some general guidelines. Like most problems in computer security, the ideal deterrents are better awareness of the issues and more openness in how systems operate in the marketplace. In particular, individual sites should clearly state the policies and methodologies they employ with recommender systems, including the role played by straddlers in their data sets and system designs. This is especially true for sites with multiple homogeneous networks (as in Figures 6c and 6d).   By conveying benefits and risks to users intuitively, recommender systems could achieve greater acceptance. We envisage three general ways to highlight the implications of our analyses. # Present plots of benefit and risk versus usermodifiable parameters such as ratings, w, and l (if the algorithm allows their direct specification) to allow users to make informed choices about their levels of involvement
PYTHIA%2DII a knowledge%2Fdatabase system for managing performance data and recommending scientific software.
this paper, we describe the PYTHIA-II architecture and its use as an RS for PDE software.
Mining and visualizing recommendation spaces for elliptic PDEs with continuous attributes.
this paper we extend previous work in mining recommendation spaces based on symbolic problem features to PDE problems with continuous-valued attributes. We identify the research issues in mining such spaces, present a dynamic programming algorithm from the data-mining literature, and describe how a priori domain metaknowledge can be used to control the complexity of induction. A visualization aid for continuous-valued recommendation spaces is also outlined. Two case studies are presented to illustrate our approach and tools: (i) a comparison of an iterative and a direct linear system solver on nearly singular problems, and (ii) a comparison of two iterative solvers on problems posed on nonrectangular domains. Both case studies involve continuously varying problem and method parameters which strongly influence the choice of best algorithm in particular cases. By mining the results from thousands of PDE solves, we can gain valuable insight into the relative performance of these methods on similar problems.
Note on generalization in experimental algorithmics.
this article, we have utilized a least-squares approximation to arrive at a more high level descriptor (described below).
PIPE Web Personalization by Partial Evaluation.
The central contribution of this paper is to model personalization by the programmatic notion of partial evaluation. Partial evaluation is a technique used to automatically specialize programs, given incomplete information about their input. The methodology presented here models a collection of information resources as a program (which abstracts the underlying schema of organization and flow of information), partially evaluates the program with respect to user input, and recreates a personalized site from the specialized program. This enables a customizable methodology called PIPE that supports the automatic specialization of resources, without enumerating the interaction sequences beforehand. Issues relating to the scalability of PIPE, information integration, sessionizing scenarios, and case studies are presented. 1
PIPE Personalizing Recommendations via Partial Evaluation

Networked Agents for Scientific Computing.
this article is that agent-based computing provides important advantages for scientific computing. We present our ideas in the context of a particular application, the simulation of gas turbine engines. This application is typical in that it involves an enormously complex device of great economic importance, one whose design is continually evolving to achieve higher value and to fit new uses.
Data Mining From Serendipity to Science %2D Guest Editors%27 Introduction.

Multidisciplinary Problem Solving Environments for Computational Science.

Multiagent System Support of Networked Scientific Computing.

