Mining Interaction Behaviors for Email Reply Order Prediction.

Context modeling for ranking and tagging bursty features in text streams.


Interval Implicitization of Parametric Surfaces.


Do You Trust to Get Trust%3F A Study of Trust Reciprocity Behaviors and Reciprocal Trust Prediction.

A Probabilistic Approach to Personalized Tag Recommendation.

Messaging Behavior Modeling in Mobile Social Networks.

TwitterRank finding topic%2Dsensitive influential twitterers.

User behavior modeling in peer%2Dto%2Dpeer file sharing networks Dissecting download and removal actions.

Modeling Syntactic Structures of Topics with a Nested HMM%2DLDA.

To Trust or Not to Trust%3F Predicting Online Trusts Using Trust Antecedent Framework.

A New Sliding%2DMode Control Algorithm with Simulation.

Methods of Time Offset Estimation and Performance Evaluation of Synchronization for 3GPP LTE Downlink.

An adaptive learning method for face hallucination using Locality Preserving Projections.

Dynamic Model and Numerical Simulation of Particle Motion in Rotation Flow Field of Centrifuge.

Knowledge transfer via multiple model local structure mapping.
The effectiveness of knowledge transfer using classification algorithms depends on the difference between the distribution that generates the training examples and the one from which test examples are to be drawn. The task can be especially difficult when the training examples are from one or several domains different from the test domain. In this paper, we propose a locally weighted ensemble framework to combine multiple models for transfer learning, where the weights are dynamically assigned according to a model’s predictive power on each test example. It can integrate the advantages of various learning algorithms and the labeled information from multiple training domains into one unified classification model, which can then be applied on a different domain. Importantly, different from many previously proposed methods, none of the base learning method is required to be specifically designed for transfer learning. We show the optimality of a locally weighted ensemble framework as a general approach to combine multiple models for domain transfer. We then propose an implementation of the local weight assignments by mapping the structures of a model onto the structures of the test domain, and then weighting each model locally according to its consistency with the neighborhood structure around the test example. Experimental results on text classification, spam filtering and intrusion detection data sets demonstrate significant improvements in classification accuracy gained by the framework. On a transfer learning task of newsgroup message categorization, the proposed locally weighted ensemble framework achieves 97 % accuracy when the best single model predicts correctly only on 73 % of the test examples. In summary, the improvement in accuracy is over 10 % and up to 30 % across different problems.
Algebraic Soft%2DDecision Decoding of Reed%2DSolomon Codes Using Bit%2DLevel Soft Information.
The performance of algebraic soft-decision decoding (ASD) of Reed-Solomon (RS) codes using bit-level soft information is investigated. Optimal multiplicity assignment strategies (MAS) of ASD with infinite cost are first studied over erasure channels and binary symmetric channels (BSC). The corresponding decoding radii are calculated in closed forms and tight bounds on the error probability are derived. The MAS and the corresponding performance analysis are then generalized to characterize the decoding region of ASD over a mixed error and bit-level erasure channel. The bit-level decoding region of the proposed MAS is shown to be significantly larger than that of conventional Berlekamp-Massey (BM) decoding. As an application, a bit-level generalized minimum distance (BGMD) decoding algorithm is proposed. The proposed BGMD compares favorably with many other RS soft-decision decoding (SDD) algorithms over various channels. Moreover, owing to the simplicity of BGMD, its performance can be tightly bounded using ordered statistics.
Instance Weighting for Domain Adaptation in NLP.
Domain adaptation is an important problem in natural language processing (NLP) due to the lack of labeled data in novel domains. In this paper, we study the domain adaptation problem from the instance weighting perspective. We formally analyze and characterize the domain adaptation problem from a distributional view, and show that there are two distinct needs for adaptation, corresponding to the different distributions of instances and classification functions in the source and the target domains. We then propose a general instance weighting framework for domain adaptation. Our empirical results on three NLP tasks show that incorporating and exploiting more information from the target domain through instance weighting is effective. 1
A two%2Dstage approach to domain adaptation for statistical classifiers.
In this paper, we consider the problem of adapting statistical classifiers trained from some source domains where labeled examples are available to a target domain where no labeled example is available. One characteristic of such a domain adaptation problem is that the examples in the source domains and the target domain are known to follow different distributions. Thus a regular classification method would tend to overfit the source domains. We present a two-stage approach to domain adaptation, where at the first generalization stage, we look for a set of features generalizable across domains, and at the second adaptation stage, we pick up useful features specific to the target domain. Observing that the exact objective function is hard to optimize, we then propose a number of heuristics to approximately achieve the goal of generalization and adaptation. Our experiments on gene name recognition using a real data set show the effectiveness of our general framework and the heuristics.
A Systematic Exploration of the Feature Space for Relation Extraction.
Relation extraction is the task of finding semantic relations between entities from text. The state-of-the-art methods for relation extraction are mostly based on statistical learning, and thus all have to deal with feature selection, which can significantly affect the classification performance. In this paper, we systematically explore a large space of features for relation extraction and evaluate the effectiveness of different feature subspaces. We present a general definition of feature spaces based on a graphic representation of relation instances, and explore three different representations of relation instances and features of different complexities within this framework. Our experiments show that using only basic unit features is generally sufficient to achieve state-of-the-art performance, while overinclusion of complex features may hurt the performance. A combination of features of different levels of complexity and from different sentence representations, coupled with task-oriented feature pruning, gives the best performance. 1
Language Models for Genomics Information Retrieval UIUC at TREC 2007 Genomics Track.

Generating gene summaries from biomedical literature A study of semi%2Dstructured summarization.

An empirical study of tokenization strategies for biomedical information retrieval.
Due to the great variation of biological names in biomedical text, appropriate tok-enization is an important preprocessing step for biomedical information retrieval. De-spite its importance, there has been little study on the evaluation of various tokeniza-tion strategies for biomedical text. In this work, we conducted a careful, systematic evaluation of a set of tokenization heuristics on all the available TREC biomedical text collections for ad hoc document retrieval, using two representative retrieval methods and a pseudo relevance feedback method. We also studied the effect of stemming and stop word removal on the retrieval performance. As expected, our experiment results show that tokenization can significantly affect the retrieval accuracy; appropriate to-kenization can improve the performance by up to 96%, measured by mean average precision (MAP). In particular, it is shown that different query types require different tokenization heuristics, stemming is effective only for certain queries, and stop word removal in general does not improve the retrieval performance on biomedical text.
A simple
In this paper we develop a simple analytic characterization of the steady state throughput, as a function of loss rate and round trip time for a bulk transfer TCP flow, i.e., a flow with an unlimited amount of data to send. Unlike the models in [6, 7, 10], our model captures not only the behavior of TCP’s fast retransmit mechanism (which is also considered in [6, 7, 10]) but also the effect of TCP’s timeout mechanism on throughput. Our measurements suggest that this latter behavior is important from a modeling perspective, as almost all of our TCP traces contained more timeout events than fast retransmit events. Our measurements demonstrate that our model is able to more accurately predict TCP throughput and is accurate over a wider range of loss rates. This material is based upon work supported by the National Science Foundation under grants NCR-95-08274, NCR-95-23807 and CDA-95-02639. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.
i%2DDBF an Improved Bloom Filter Representation Method on Dynamic Set.

Exploiting Domain Structure for Named Entity Recognition.
Named Entity Recognition (NER) is a fundamental task in text mining and natural language understanding. Current approaches to NER (mostly based on supervised learning) perform well on domains similar to the training domain, but they tend to adapt poorly to slightly different domains. We present several strategies for exploiting the domain structure in the training data to learn a more robust named entity recognizer that can perform well on a new domain. First, we propose a simple yet effective way to automatically rank features based on their generalizabilities across domains. We then train a classifier with strong emphasis on the most generalizable features. This emphasis is imposed by putting a rank-based prior on a logistic regression model. We further propose a domain-aware cross validation strategy to help choose an appropriate parameter for the rank-based prior. We evaluated the proposed method with a task of recognizing named entities (genes) in biology text involving three species. The experiment results show that the new domainaware approach outperforms a state-ofthe-art baseline method in adapting to new domains, especially when there is a great difference between the new domain and the training domain.
Automatically Generating Gene Summaries from Biomedical Literature.
Biologists often need to find information about genes whose function is not described in the genome databases. Currently they must try to search disparate biomedical literature to locate relevant articles, and spend considerable efforts reading the retrieved articles in order to locate the most relevant knowledge about the gene. We describe our software, the first that automatically generates gene summaries from biomedical literature. We present a two-stage summarization method, which involves first retrieving relevant articles and then extracting the most informative sentences from the retrieved articles to generate a structured gene summary. The generated summary explicitly covers multiple aspects of a gene, such as the sequence information, mutant phenotypes, and molecular interaction with other genes. We propose several heuristic approaches to improve the accuracy in both stages. The proposed methods are evaluated using 10 randomly chosen genes from FlyBase and a subset of Medline abstracts about Drosophila. The results show that the precision of the top selected sentences in the 6 aspects is typically about 50-70%, and the generated summaries are quite informative, indicating that our approaches are effective in automatically summarizing literature information about genes. The generated summaries not only are directly useful to biologists but also serve as useful entry points to enable them to quickly digest the retrieved literature articles. 1.
Robust Pseudo Feedback Estimation and HMM Passage Extraction UIUC at TREC 2006 Genomics Track.

Extraction of coherent relevant passages using hidden Markov models.

Multilevel Coding for Channels with Non%2Duniform Inputs and Rateless Transmission over the BSC
We consider coding schemes for channels with nonuniform inputs (NUI), where standard linear block codes can not be applied directly. We show that multilevel coding (MLC) with a set of linear codes and a deterministic mapper can achieve the information rate of the channel with NUI. The mapper, however, does not have to be one-to-one. As an application of the proposed MLC scheme, we present a rateless transmission scheme over the binary symmetric channel (BSC).
Algebraic Soft%2DDecision Decoding of Reed%2DSolomon Codes Using Bit%2Dlevel Soft Information
The performance of algebraic soft-decision decoding (ASD) of Reed-Solomon (RS) codes using bit-level soft information is investigated. Optimal multiplicity assignment strategies (MAS) of ASD with infinite cost are first studied over erasure channels and binary symmetric channels (BSC). The corresponding decoding radii are calculated in closed forms and tight bounds on the error probability are derived. The MAS and the corresponding performance analysis are then generalized to characterize the decoding region of ASD over a mixed error and bit-level erasure channel. The bit-level decoding region of the proposed MAS is shown to be significantly larger than that of conventional Berlekamp-Massey (BM) decoding. As an application, a bit-level generalized minimum distance (BGMD) decoding algorithm is proposed. The proposed BGMD compares favorably with many other RS soft-decision decoding (SDD) algorithms over various channels. Moreover, owing to the simplicity of BGMD, its performance can be tightly bounded using ordered statistics.
Greedy scheduling performance for a zero%2Dforcing dirty%2Dpaper coded system.


Accurately extracting coherent relevant passages using hidden Markov models.


Performance Analysis of Algebraic Soft Decoding of Reed%2DSolomon Codes over Binary Symmetric and Erasure Channels
Abstract — In this paper, we characterize the decoding region of algebraic soft decoding (ASD) [4] of Reed-Solomon (RS) codes over some discrete alphabet channels (DAC). Optimal multiplicity assignment strategies (MAS) are investigated and tight bounds are derived to show the performance of ASD over binary erasure channel (BEC) and binary symmetric channel (BSC). The performance analysis suggests that ASD can provide a significant gain over conventional Berlekamp Massey (BM) decoding. The analysis technique can also be extended to other channel models, e.g., RS coded modulation over erasure channels. The results suggest potential application of ASD to practical systems [7]. I.
Iterative Soft Input Soft Output Decoding of Reed%2DSolomon Codes by Adapting the Parity Check Matrix
Abstract — An iterative algorithm is presented for soft-inputsoft-output (SISO) decoding of Reed-Solomon (RS) codes. The proposed iterative algorithm uses the sum product algorithm (SPA) in conjunction with a binary parity check matrix of the RS code. The novelty is in reducing a submatrix of the binary parity check matrix that corresponds to less reliable bits to a sparse nature before the SPA is applied at each iteration. The proposed algorithm can be geometrically interpreted as a twostage gradient descent with an adaptive potential function. This adaptive procedure is crucial to the convergence behavior of the gradient descent algorithm and, therefore, significantly improves the performance. Simulation results show that the proposed decoding algorithm and its variations provide significant gain over hard decision decoding (HDD) and compare favorably with other popular soft decision decoding methods. Index Terms — adapting the parity check matrix, gradient descent, iterative decoding, soft decision decoding, Reed-Solomon (RS) codes. I.
A New Method to Measure Cross Sectional Area of Vessels in MRI Image and Its Application in Stenosis Detection.

UIUC in HARD 2004%2D%2DPassage Retrieval Using HMMs.

Antenna diversity in multiuser data networks.

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.
Towards integrated PSEs for wireless communications experiences with the S4W and SitePlanner%AE projects.

Adaptive Filters for Continuous Queries over Distributed Data Streams.
We consider an environment where distributed  data sources continuously stream updates to a  centralized processor that monitors continuous  queries over the distributed data. Significant communication  overhead is incurred in the presence of  rapid update streams, and we propose a new technique  for reducing the overhead. Users register  continuous queries with precision requirements at  the central stream processor, which installs filters  at remote data sources. The filters adapt to changing  conditions to minimize stream rates while  guaranteeing that all continuous queries still receive  the updates necessary to provide answers of  adequate precision at all times. Our approach enables  applications to trade precision for communication  overhead at a fine granularity by individually  adjusting the precision constraints of continuous  queries over streams in a multi-query workload.
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
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.
BSML A Binding Schema Markup Language for Data Interchange in Problem Solving Environments %28PSEs%29

