Yong Zeng

Breaking out of the MisMatch trap

When users issue a query to a database, they have expectations about the results. If what they search for is unavailable in the database, the system will return an empty result or, worse, erroneous mismatch results.We call this problem the MisMatch Problem. In this paper, we solve the MisMatch problem in the context of XML keyword search. Our solution is based on two novel concepts that we introduce: Target Node Type and Distinguishability. Using these concepts, we develop a low-cost post-processing algorithm on the results of query evaluation to detect the MisMatch problem and generate helpful suggestions to users. Our approach has three noteworthy features: (1) for queries with the MisMatch problem, it generates the explanation, suggested queries and their sample results as the output to users, helping users judge whether the MisMatch problem is solved without reading all query results; (2) it is portable as it can work with any LCA-based matching semantics and is orthogonal to the choice of result retrieval method adopted; (3) it is lightweight in the way that it occupies a very small proportion of the whole query evaluation time. Extensive experiments on three real datasets verify the effectiveness, efficiency and scalability of our approach. A search engine called XClear has been built and is available at http://xclear.comp.nus.edu.sg.

A general framework to resolve the MisMatch problem in XML keyword search

When users issue a query to a database, they have expectations about the results. If what they search for is unavailable in the database, the system will return an empty result or, worse, erroneous mismatch results. We call this problem the MisMatch problem. In this paper, we solve the MisMatch problem in the context of XML keyword search. Our solution is based on two novel concepts that we introduce: target node type and Distinguishability. Target Node Type represents the type of node a query result intends to match, and Distinguishability is used to measure the importance of the query keywords. Using these concepts, we develop a low-cost post-processing algorithm on the results of query evaluation to detect the MisMatch problem and generate helpful suggestions to users. Our approach has three noteworthy features: (1) for queries with the MisMatch problem, it generates the explanation, suggested queries and their sample results as the output to users, helping users judge whether the MisMatch problem is solved without reading all query results; (2) it is portable as it can work with any lowest common ancestor-based matching semantics (for XML data without ID references) or minimal Steiner tree-based matching semantics (for XML data with ID references) which return tree structures as results. It is orthogonal to the choice of result retrieval method adopted; (3) it is lightweight in the way that it occupies a very small proportion of the whole query evaluation time. Extensive experiments on three real datasets verify the effectiveness, efficiency and scalability of our approach. A search engine called XClear has been built and is available at http://xclear.comp.nus.edu.sg.

Exploratory Keyword Search with Interactive Input

Due to the intrinsic ambiguity of keyword queries, users usually need to reformulate their queries multiple times to get the desired information. Even worse, users either have no way to precisely specify their search intention, or have limited domain knowledge on the data to precisely express their search intention. Moreover, they may just have a general interest to explore the data by keyword query. Therefore, our goal is to design an exploratory search paradigm that is able to bring humans more actively into the search process, in order to meet various user information needs, ranging from simple lookup to learning and understanding of the data. Besides, keyword queries against data with structure, such as XML, can run into multiple difficulties: how to identify the search target; more types of ambiguity arise as a keyword can be part of the structure as well as content of data, etc. Effectively addressing these requires solutions to multiple challenges. While some have been addressed to some extent individually, there is no previous effort to develop a comprehensive system to meet these important user needs and meet all of these challenges. Therefore, we propose a framework called ClearMap that natively supports visualized exploratory search paradigm on XML data. In particular, we offer an interactive and visualized mechanism to present the outcome of the query, enable user to explore and manipulate the underlying data to either quickly find desired information or learn the relationship among data items, as well as provide interactive suggestions when their expected results do not exist in the data. A preliminary version of ClearMap and its source code are available for try at http://xmlclearmap.comp.nus.edu.sg.

Removing the mismatch headache in XML keyword search

In this demo, we study one category of query refinement problems in the context of XML keyword search, where what users search for do not exist in the data while useless results are returned by the search engine. It is a hidden but important problem. We refer to it as the MisMatch problem. We propose a practical yet efficient way to detect the MisMatch problem and generate helpful suggestions to users, namely MisMatch detector and suggester. Our approach can be viewed as a post-processing job of query evaluation. An online XML keyword search engine embedding the MisMatch detector and suggester has been built and is available at [1].

Efficient XML Keyword Search: From Graph Model to Tree Model

Keyword search, as opposed to traditional structured query, has been becoming more and more popular on querying XML data in recent years. XML documents usually contain some ID nodes and IDREF nodes to represent reference relationships among the data. An XML document with ID/IDREF is modeled as a graph by existing works, where the keyword query results are computed by graph traversal. As a comparison, if ID/IDREF is not considered, an XML document can be modeled as a tree. Keyword search on XML tree can be much more efficient using tree-based labeling techniques. A nature question is whether we need to abandon the efficient XML tree search methods and invent new, but less efficient search methods for XML graph. To address this problem, we propose a novel method to transform an XML graph to a tree model such that we can exploit existing XML tree search methods. The experimental results show that our solution can outperform the traditional XML graph search methods by orders of magnitude in efficiency while generating a similar set of results as existing XML graph search methods.

A Hybrid Approach for General XML Query Processing

The state-of-the-art XML twig pattern query processing algorithms focus on matching a single twig pattern to a document. However, many practical queries are modeled by multiple twig patterns with joins to link them. The output of twig pattern matching is tuples of labels, while the joins between twig patterns are based on values. The inefficiency of integrating label-based structural joins in twig pattern matching and value-based joins to link patterns becomes an obstacle preventing those structural join algorithms in literatures from being adopted in practical XML query processors. In this paper, we propose a hybrid approach to bridge this gap. In particular, we introduce both relational tables and inverted lists to organize values and elements respectively. General XML queries involving several twig patterns are processed by the both data structures. We further analyze join order selection for a general query with both pattern matching and value-based join, which is essential for the generation of a good query plan.

Supporting range queries in XML keyword search

XML data is normally queried by rigorous structured query languages, e.g., XPath, XQuery, etc. In recent years keyword search has become more and more popular because it provides a more user-friendly way to explore data. Keyword search on XML data has also been a hot research issue recently. So far none of the existing XML keyword search methods has considered range queries. In this paper we point out that supporting range queries in XML keyword search is beneficial and non-trivial to the user, especially in the case of querying business semi-structured data, where numerals (like stock price, product quantity, market share percentage, etc.) could be the main part of the data. Actually existing XML keyword search methods do not support range queries at two levels: keyword query syntax level and keyword search method level. To support range queries in XML keyword search: (1) we enrich the current XML keyword query syntax to let the user make range specification; (2) we then extend existing XML keyword search methods by proposing a new index to support both range match and point match. The new index is transparent to existing XML keyword search methods. It can seamlessly work with them and well support range queries in XML keyword search.