Query portals dynamically generating portals for entity%2Doriented web queries.

Keyword   A Framework to Improve Keyword Search Over Entity Databases.
Keyword search over entity databases (e.g., product, movie databases) is an important problem. Current techniques for keyword search on databases may often return incomplete and imprecise results. On the one hand, they either require that relevant entities contain all (or most) of the query keywords, or that relevant entities and the query keywords occur together in several documents from a known collection. Neither of these requirements may be satisfied for a number of user queries. Hence results for such queries are likely to be incomplete in that highly relevant entities may not be returned. On the other hand, although some returned entities contain all (or most) of the query keywords, the intention of the keywords in the query could be different from that in the entities. Therefore, the results could also be imprecise. To remedy this problem, in this paper, we propose a general framework that can improve an existing search interface by translating a keyword query to a structured query. Specifically, we leverage the keyword to attribute value associations discovered in the results returned by the original search interface. We show empirically that the translated structured queries alleviate the above problems. 1.
Exploiting web search to generate synonyms for entities.

Exploiting web search engines to search structured databases.

Detecting gene clusters under evolutionary constraint in a large number of genomes.

Top%2Ddown mining of frequent closed patterns from very high dimensional data.

Promotion Analysis in Multi%2DDimensional Space.
Promotion is one of the key ingredients in marketing. It is often desirable to find merit in an object (e.g., product, person, organization, or service) and promote it in an appropriate community. In this paper, we propose a novel functionality, called promotion analysis through ranking, for promoting a given object by leveraging highly ranked results. Since the object may not be highly ranked in the global space, our goal is to discover promotive subspaces in which the object becomes prominent. To achieve this goal, the notion of promotiveness is formulated. We show that this functionality is practical and useful in a wide variety of applications such as business intelligence. However, computing promotive subspaces is challenging due to the explosion of search space and high aggregation cost. For efficient computation, we propose a PromoRank framework, and develop three efficient optimization techniques, namely subspace pruning, object pruning, and promotion cube, which are seamlessly integrated into the framework. Our empirical evaluation on two real data sets confirms the effectiveness of promotion analysis, and that our proposed algorithms significantly outperform baseline solutions. 1.
Mining Document Collections to Facilitate Accurate Approximate Entity Matching.
Many entity extraction techniques leverage large reference entity tables to identify entities in documents. Often, an entity is referenced in document collections differently from that in the reference entity tables. Therefore, we study the problem of determining whether or not a substring “approximately” matches with a reference entity. Similarity measures which exploit the correlation between candidate substrings and reference entities across a large number of documents are known to be more robust than traditional stand alone string-based similarity functions. However, such an approach has significant efficiency challenges. In this paper, we adopt a new architecture and propose new techniques to address these efficiency challenges. We mine document collections and expand a given reference entity table with variations of each of its entities. Thus, the problem of approximately matching an input string against reference entities reduces to that of exact match against the expanded reference table, which can be implemented efficiently. In an extensive experimental evaluation, we demonstrate the accuracy and scalability of our techniques. 1.
Mining Document Collections to Facilitate Accurate Approximate Entity Matching.
Many entity extraction techniques leverage large reference entity tables to identify entities in documents. Often, an entity is referenced in document collections differently from that in the reference entity tables. Therefore, we study the problem of determining whether or not a substring “approximately” matches with a reference entity. Similarity measures which exploit the correlation between candidate substrings and reference entities across a large number of documents are known to be more robust than traditional stand alone string-based similarity functions. However, such an approach has significant efficiency challenges. In this paper, we adopt a new architecture and propose new techniques to address these efficiency challenges. We mine document collections and expand a given reference entity table with variations of each of its entities. Thus, the problem of approximately matching an input string against reference entities reduces to that of exact match against the expanded reference table, which can be implemented efficiently. In an extensive experimental evaluation, we demonstrate the accuracy and scalability of our techniques. 1.
P%2DCube Answering Preference Queries in Multi%2DDimensional Space.
 Many new applications that involve decision making need online (i.e., OLAP-styled) preference analysis with multidimensional boolean selections. Typical preference queries includes top-k queries and skyline queries. An analytical query often comes with a set of boolean predicates that constrain a target subset of data, which, may also vary incrementally by drilling/rolling operators. To efficiently support preference queries with multiple boolean predicates, neither boolean-thenpreference nor preference-then-boolean approach is satisfactory. To integrate boolean pruning and preference pruning in a unified framework, we propose signature, a new materialization measure for multi-dimensional group-bys. Based on this, we propose P-Cube (i.e., data cube for preference queries) and study its complete life cycle, including signature generation, compression, decomposition, incremental maintenance and usage for efficient on-line analytical query processing. We present a signature-based progressive algorithm that is able to simultaneously push boolean and preference constraints deep into the database search. Our performance study shows that the proposed method achieves at least one order of magnitude speed-up over existing approaches.  
ARCube supporting ranking aggregate queries in partially materialized data cubes.
Supporting ranking queries in database systems has been a popular research topic recently. However, there is a lack of study on supporting ranking queries in data warehouses where ranking is on multidimensional aggregates instead of on measures of base facts. To address this problem, we propose a query execution model to answer different types of ranking aggregate queries based on a unified, partial cube structure, ARCube. The query execution model follows a candidate generation and verification framework, where the most promising candidate cells are generated using a set of high-level guiding cells. We also identify a bounding principle for effective pruning: once a guiding cell is pruned, all of its children candidate cells can be pruned. We further address the problem of efficient online candidate aggregation and verification by developing a chunk-based execution model to verify a bulk of candidates within a bounded memory buffer. Our extensive performance study shows that the new framework not only leads to an order of magnitude performance improvements over the state-of-the-art method, but also is much more flexible in terms of the types of ranking aggregate queries supported.
An efficient filter for approximate membership checking.

Efficiently Identifying Max%2DGap Clusters in Pairwise Genome Comparison.

Integrating OLAP and Ranking The Ranking%2DCube Methodology.

Progressive and selective merge computing top%2Dk with ad%2Dhoc ranking functions.
The family of threshold algorithm (i.e., TA) has been widely studied for efficiently computing top-k queries. TA uses a sort-merge framework that assumes data lists are pre-sorted, and the ranking functions are monotone. However, in many database applications, attribute values are indexed by treestructured indices (e.g., B-tree, R-tree), and the ranking functions are not necessarily monotone. To answer top-k queries with ad-hoc ranking functions, this paper studies an index-merge paradigm that performs progressive search over the space of joint states composed by multiple index nodes. We address two challenges for efficient query processing. First, to minimize the search complexity, we present a doubleheap algorithm which supports not only progressive state search but also progressive state generation. Second, to avoid unnecessary disk access, we characterize a type of “empty-state ” that does not contribute to the final results, and propose a new materialization model, join-signature, to prune empty-states. Our performance study shows that the proposed method achieves one order of magnitude speed-up over baseline solutions.
DataScope Viewing Database Contents in Google Maps%27 Way.

On compressing frequent patterns.
A major challenge in frequent-pattern mining is the sheer size of its mining results. To compress the frequent patterns, we propose to cluster frequent patterns with a tightness measure δ (called δ-cluster), and select a representative pattern for each cluster. The problem of finding a minimum set of representative patterns is shown NP-Hard. We develop two greedy methods, RPglobal and RPlocal. The former has the guaranteed compression bound but higher computational complexity. The latter sacrifices the theoretical bounds but is far more efficient. Our performance study shows that the compression quality using RPlocal is very close to RPglobal, and both can reduce the number of closed frequent patterns by almost two orders of magnitude. Furthermore, RPlocal mines even faster than FPClose[10], a very fast closed frequent-pattern mining method. We also show that RPglobal and RPlocal can be combined together to balance the quality and efficiency.
Frequent pattern mining current status and future directions.

Computing Iceberg Cubes by Top%2DDown and Bottom%2DUp Integration The StarCubing Approach.

Optimization of Bounds in Temporal Flexible Plans with Dynamic Controllability.

Semantic annotation of frequent patterns.
Using frequent patterns to analyze data has been one of the fundamental approaches in many data mining applications. Research in frequent pattern mining has so far mostly focused on developing efficient algorithms to discover various kinds of frequent patterns, but little attention has been paid to the important next step—interpreting the discovered frequent patterns. Although the compression and summarization of frequent patterns has been studied in some recent work, the proposed techniques there can only annotate a frequent pattern with nonsemantical information (e.g., support), which provides only limited help for a user to understand the patterns. In this article, we study the novel problem of generating semantic annotations for frequent patterns. The goal is to discover the hidden meanings of a frequent pattern by annotating it with in-depth, concise, and structured information. We propose a general approach to generate such an annotation for a frequent pattern by constructing its context model, selecting informative context indicators, and extracting representative transactions and semantically similar patterns. This general approach can well incorporate the user’s prior knowledge, and has potentially many applications, such as generating a dictionary-like description for a pattern, finding synonym patterns, discovering semantic relations, and summarizing semantic classes of a set of frequent patterns. Experiments
Top%2DDown Mining of Interesting Patterns from Very High Dimensional Data.

C%2DCubing Efficient Computation of Closed Cubes by Aggregation%2DBased Checking.
It is well recognized that data cubing often produces huge outputs. Two popular efforts devoted to this problem are (1) iceberg cube, where only significant cells are kept, and (2) closed cube, where a group of cells which preserve roll-up/drill-down semantics are losslessly compressed to one cell. Due to its usability and importance, efficient computation of closed cubes still warrants a thorough study. In this paper, we propose a new measure, called closedness, for efficient closed data cubing. We show that closedness is an algebraic measure and can be computed efficiently and incrementally. Based on closedness measure, we develop an an aggregation-based approach, called C-Cubing (i.e., Closed-Cubing), and integrate it into two successful iceberg cubing algorithms: MM-Cubing and Star-Cubing. Our performance study shows that C-Cubing runs almost one order of magnitude faster than the previous approaches. We further study how the performance of the alternative algorithms of C-Cubing varies w.r.t the properties of the data sets. 1
Generating semantic annotations for frequent patterns with context analysis.
As a fundamental data mining task, frequent pattern mining has widespread applications in many different domains. Research in frequent pattern mining has so far mostly focused on developing efficient algorithms to discover various kinds of frequent patterns, but little attention has been paid to the important next step – interpreting the discovered frequent patterns. Although some recent work has studied the compression and summarization of frequent patterns, the proposed techniques can only annotate a frequent pattern with non-semantical information (e.g. support), which provides only limited help for a user to understand the patterns. In this paper, we propose the novel problem of generating semantic annotations for frequent patterns. The goal is to annotate a frequent pattern with in-depth, concise, and structured information that can better indicate the hidden meanings of the pattern. We propose a general approach to generate such an annotation for a frequent pattern by constructing its context model, selecting informative context indicators, and extracting representative transactions and semantically similar patterns. This general approach has potentially many applications such as generating a dictionarylike description for a pattern, finding synonym patterns, discovering semantic relations, and summarizing semantic classes of a set of frequent patterns. Experiments on different datasets show that our approach is effective in generating semantic pattern annotations.
Extracting redundancy%2Daware top%2Dk patterns.

Mining Interesting Patterns from Very High Dimensional Data A Top%2DDown Row Enumeration Approach.

Ranking objects based on relationships.
Text corpora are often enhanced by additional metadata which relate real-world entities, with each document in which such entities are discussed. Such relationships are typically obtained through widely available Information Extraction tools. At the same time, interesting known associations typically hold among these entities. For instance, a corpus might contain discussions on hotels, cities and airlines; fixed associations among these entities may include: airline A operates a flight to city C, hotel H is located in city C. A plethora of applications necessitate the identification of associated entities, each best matching a given set of keywords. Consider the sample query: Find a holiday package in a “pet-friendly ” hotel, located in a “historical ” yet “lively ” city, with travel operated by an “economical ” and “safe ” airline. These keywords are unlikely to occur in the textual description of entities themselves, (e.g., the actual hotel name or the city name or the airline name). Consequently to answer such queries, one needs to exploit both relationships between entities and documents (e.g., keyword “pet-friendly ” occurs in a document that contains an entity specifying a hotel name H), and the known associations between entities (e.g., hotel H is located in city C). In this work, we focus on the class of “entity package finder” queries outlined above. We demonstrate that existing techniques cannot be efficiently adapted to solve this problem, as the resulting algorithm relies on estimations with excessive runtime and/or storage overheads. We propose an efficient algorithm to process such queries, over large corpora. We devise early pruning and termination strategies, in the presence of joins and aggregations (executed on entities extracted from text), that do not depend on any estimates. Our analysis and experimental evaluation on real and synthetic data demonstrates the efficiency and scalability of our approach. 1.
Towards Robust Indexing for Ranked Queries.
RESEARCH INTERESTS I am generally interested in data mining, data warehousing and database systems. My current research focuses on modeling, managing and analyzing large-scale graph and information network data.
Answering Top%2Dk Queries with Multi%2DDimensional Selections The Ranking Cube Approach.
Observed in many real applications, a top-k query often consists of two components to reflect a user’s preference: a selection condition and a ranking function. A user may not only propose ad hoc ranking functions, but also use different interesting subsets of the data. In many cases, a user may want to have a thorough study of the data by initiating a multi-dimensional analysis of the top-k query results. Previous work on top-k query processing mainly focuses on optimizing data access according to the ranking function only. The problem of efficient answering top-k queries with multidimensional selections has not been well addressed yet. This paper proposes a new computational model, called ranking cube, for efficient answering top-k queries with multidimensional selections. We define a rank-aware measure for the cube, capturing our goal of responding to multidimensional ranking analysis. Based on the ranking cube, an efficient query algorithm is developed which progressively retrieves data blocks until the top-k results are found. The curse of dimensionality is a well-known challenge for the data cube and we cope with this difficulty by introducing a new technique of ranking fragments. Our experiments on Microsoft’s SQL Server 2005 show that our proposed approaches have significant improvement over the previous methods. 1.
Answering Top%2Dk Queries with Multi%2DDimensional Selections The Ranking Cube Approach.
Observed in many real applications, a top-k query often consists of two components to reflect a user’s preference: a selection condition and a ranking function. A user may not only propose ad hoc ranking functions, but also use different interesting subsets of the data. In many cases, a user may want to have a thorough study of the data by initiating a multi-dimensional analysis of the top-k query results. Previous work on top-k query processing mainly focuses on optimizing data access according to the ranking function only. The problem of efficient answering top-k queries with multidimensional selections has not been well addressed yet. This paper proposes a new computational model, called ranking cube, for efficient answering top-k queries with multidimensional selections. We define a rank-aware measure for the cube, capturing our goal of responding to multidimensional ranking analysis. Based on the ranking cube, an efficient query algorithm is developed which progressively retrieves data blocks until the top-k results are found. The curse of dimensionality is a well-known challenge for the data cube and we cope with this difficulty by introducing a new technique of ranking fragments. Our experiments on Microsoft’s SQL Server 2005 show that our proposed approaches have significant improvement over the previous methods. 1.
Mining Evolving Customer%2DProduct Relationships in Multi%2DDimensional Space.
Previous work on mining transactional database has focused primarily on mining frequent itemsets, association rules, and sequential patterns. However, interesting relationships between customers and items, especially their evolution with time, have not been studied thoroughly. In this paper, we propose a Gaussin transformation-based regression model that captures time-variant relationships between customers and products. Moreover, since it is interesting to discover such relationships in a multi-dimensional space, an efficient method has been developed to compute multi-dimensional aggregates of such curves in a data cube environment. Our experimental results have demonstrated the promise of the approach.
Summarizing itemset patterns a profile%2Dbased approach.
Frequent-pattern mining has been studied extensively on scalable methods for mining various kinds of patterns including itemsets, sequences, and graphs. However, the bottleneck of frequent-pattern mining is not at the efficiency but at the interpretability, due to the huge number of patterns generated by the mining process. In this paper, we examine how to summarize a collection of itemset patterns using only K representatives, a small number of patterns that a user can handle easily. The K representatives should not only cover most of the frequent patterns but also approximate their supports. A generative model is built to extract and profile these representatives, under which the supports of the patterns can be easily recovered without consulting the original dataset. Based on the restoration error, we propose a quality measure function to determine the optimal value of parameter K. Polynomial time algorithms are developed together with several optimization heuristics for efficiency improvement. Empirical studies indicate that we can obtain compact summarization in real datasets.
Mining Compressed Frequent%2DPattern Sets.
A major challenge in frequent-pattern mining is the sheer size of its mining results. In many cases, a high min sup threshold may discover only commonsense patterns but a low one may generate an explosive number of output patterns, which severely restricts its usage. In this paper, we study the problem of compressing frequent-pattern sets. Typically, frequent patterns can be clustered with a tightness measure δ (called δ-cluster), and a representative pattern can be selected for each cluster. Unfortunately, finding a minimum set of representative patterns is NP-Hard. We develop two greedy methods, RPglobal and RPlocal. The former has the guaranteed compression bound but higher computational complexity. The latter sacrifices the theoretical bounds but is far more efficient. Our performance study shows that the compression quality using RPlocal is very close to RPglobal, and both can reduce the number of closed frequent patterns by almost two orders of magnitude. Furthermore, RPlocal mines even faster than FPClose[11], a very fast closed frequentpattern mining method. We also show that RPglobal and RPlocal can be combined together to balance the quality and efficiency. 1
Optimization of Bounds in Temporal Flexible Planning with Dynamic Controllability.

MM%2DCubing Computing Iceberg Cubes by Factorizing the Lattice Space.
The data cube and iceberg cube computation problem has been studied by many researchers. There are three major approaches developed in this direction: (1) top-down computation, represented by MultiWay array aggregation [22], which utilizes shared computation and performs well on dense data sets; (2) bottom-up computation, represented by BUC [6], which takes advantage of Apriori Pruning and performs well on sparse data sets; and (3) integrated top-down and bottom-up computation, represented by StarCubing [21], which takes advantages of both and has high performance in most cases. However, the performance of Star-Cubing degrades in very sparse data sets due to the additional cost introduced by the tree structure. None of the three approaches achieves uniformly high performance on all kinds of data sets.
Star%2DCubing Computing Iceberg Cubes by Top%2DDown and Bottom%2DUp Integration.
Data cube computation is one of the most essential but expensive operations in data warehousing. Previous studies have developedtwo major approaches, top-down vs. bottomup. The former, represented by the MultiWay Array Cube (called MultiWay) algorithm[25], aggregates simultaneously on multipledimensions
