Catálogo de publicaciones - libros

We consider the retrieval of XML-structured documents, and of passages from such documents, defined as elements of the XML structure. These are considered from the point of view of passage retrieval, as a form of document retrieval. A retrievable unit (an element chosen as defining suitable passages for retrieval) is a textual document in its own right, but may inherit information from the other parts of the same document. Again, this inheritance is defined in terms of the XML structure. All retrievable units are mapped onto a common field structure, and the ranking function is a standard document retrieval function with a suitable field weighting. A small experiment to demonstrate the idea, using INEX data, is described.

Converting a natural language query sentence into a formal database query is a major challenge. We have constructed NaLIX, a natural language interface for querying XML data. Through our experience with NaLIX, we find that failures in natural language query understanding can often be dealt with as ambiguities in term meanings. These failures are typically the result of either the user’s poor knowledge of the database schema or the system’s lack of linguistic coverage. With automatic term expansion techniques and appropriate interactive feedback, we are able to resolve these ambiguities. In this paper, we describe our approach and present results demonstrating its effectiveness.

In XML retrieval, information retrieval systems have to find out which are the most appropriate retrieval units and return only these to the user, avoiding overlapping elements in the result lists. This paper studies structural relationships between elements and explains how they can be used to produce a better ranking for a focused task. We analise relevance judgements to find the most useful links between elements and show how a retrieval model can be adapted to incorporate this information. Experiments on the INEX 2005 test collection show that the structural relationships improve retrieval effectiveness considerably.

This paper proposes a method of ranking XML documents with respect to an Information Retrieval query by means of fuzzy logic. The proposed method allows imprecise queries to be evaluated against an XML document collection and it provides a model of ranking XML documents. In addition the proposed method enables sophisticated ranking of documents by employing proximity measures and the concept of editing (Levenshtein) distance between terms or XML paths.

In this paper we present a novel news filtering model based on flexible and soft filtering criteria and exploiting a fuzzy hierarchical categorization of news. The filtering module is designed to provide news professionals and general users with an interactive and personalised tool for news gathering and delivery. It exploits content-based filtering criteria and category-based filtering techniques to deliver to the user a ranked list of either news or clusters of news. In fact, if the user prefers to have a synthetic view of the topics of recent news pushed by the stream, the system filters groups (clusters) of news having homogenous contents, identified automatically by the application of a fuzzy clustering algorithm that organizes the recent news into a fuzzy hierarchy. The filter can be trained explicitly by the user to learn his/her interests as well as implicitly by monitoring his/her interaction with the system. Several filtering criteria can be applied to select and rank news to the users based on the user’s information preferences and presentation preferences. User preferences specify what information (the contents of interest) is relevant to the user, the sources that provide reliable information, and the period of time during which the information remains relevant. Each individual news or cluster of news homogeneous with respect to their content is selected based on a customizable multi criteria decision making approach and ranked based on a combination of criteria specified by the user in his/her presentation preferences.

Searchers’ difficulty in formulating effective queries for their information needs is well known. Analysis of search session logs shows that users often pose short, vague queries and then struggle with revising them. Interactive query expansion (users selecting terms to add to their queries) dramatically improves effectiveness and satisfaction. Suggesting relevant candidate expansion terms based on the initial query enables users to satisfy their information needs faster. We find that suggesting query phrases other users have found it necessary to add for a given query (mined from session logs) dramatically improves the quality of suggestions over simply using cooccurrence. However, this exacerbates the sparseness problem faced when mining short queries that lack features. To mitigate this, we tag query phrases with higher level topical categories to mine more general rules, finding that this enables us to make suggestions for approximately 10% more queries while maintaining an acceptable false positive rate.

When querying XML collections, users cannot always express their need in a precise way. Systems should therefore support vagueness at both the content and structural level of queries. This paper present a relevance-oriented method for ranking XML components. The aim here is to evaluate whether structural hints help to better answer the user needs. We experiment (within the INEX framework) with users needs expressed in a flexible way (i.e with ou without structural hints). Results show that they clearly improve performance, even if they are expressed in an ”artificial way”. Relevance seems therefore to be closely linked to structure. Moreover, too complex structural hints do not lead to better results.

XML has become a widespread format for data exchange over the Internet. The current state of the art in querying XML data is represented by XPath and XQuery, both of which define binary predicates. In this paper, we advocate that binary selection can at times be restrictive due to very nature of XML, and to the uses that are made of it. We therefore suggest a querying framework, called FXPath, based on fuzzy logics. In particular, we propose the use of fuzzy predicates for the definition of more “vague” and softer queries. We also introduce a function called “deep-similar”, which aims at substituting XPath’s typical “deep-equal” function. Its goal is to provide a degree of similarity between two XML trees, assessing whether they are similar both structure-wise and content-wise. The approach is exemplified in the field of e-learning metadata.

This paper describes a flexible information retrieval approach based on CP-Nets (Conditional Preferences Networks). The CP-Net formalism is used for both representing qualitative queries (expressing user preferences) and representing documents in order to carry out the retrieval process. Our contribution focuses on the difficult task of term weighting in the case of qualitative queries. In this context, we propose an accurate algorithm based on UCP-Net features to automatically weight Boolean queries. Furthermore, we also propose a flexible approach for query evaluation based on a flexible aggregation operator adapted to the CP-Net semantics.

Highly heterogeneous XML collections are thematic collections exploiting different structures: the parent-child or ancestor-descendant relationships are not preserved and vocabulary discrepancies in the element names can occur. In this setting current approaches return answers with low precision. By means of similarity measures and semantic inverted indices we present an approach for improving the precision of query answers without compromising performance.