Structured queries,language modeling,and relevance modeling in cross-language information retrieval

Two probabilistic approaches to cross-lingual retrieval are in wide use today, those based on probabilistic models of relevance, as exemplified by INQUERY, and those based on language modeling. INQUERY, as a query net model, allows the easy incorporation of query operators, including a synonym operator, which has proven to be extremely useful in cross-language information retrieval (CLIR), in an approach often called structured query translation. In contrast, language models incorporate translation probabilities into a unified framework. We compare the two approaches on Arabic and Spanish data sets, using two kinds of bilingual dictionaries––one derived from a conventional dictionary, and one derived from a parallel corpus. We find that structured query processing gives slightly better results when queries are not expanded. On the other hand, when queries are expanded, language modeling gives better results, but only when using a probabilistic dictionary derived from a parallel corpus.We pursue two additional issues inherent in the comparison of structured query processing with language modeling. The first concerns query expansion, and the second is the role of translation probabilities. We compare conventional expansion techniques (pseudo-relevance feedback) with relevance modeling, a new IR approach which fits into the formal framework of language modeling. We find that relevance modeling and pseudo-relevance feedback achieve comparable levels of retrieval and that good translation probabilities confer a small but significant advantage.     

Improving probabilistic information retrieval by modeling burstiness of words

The classical probabilistic models attempt to capture the ad hoc information retrieval problem within a rigorous probabilistic framework. It has long been recognized that the primary obstacle to the effective performance of the probabilistic models is the need to estimate a relevance model. The Dirichlet compound multinomial (DCM) distribution based on the Polya Urn scheme, which can also be considered as a hierarchical Bayesian model, is a more appropriate generative model than the traditional multinomial distribution for text documents. We explore a new probabilistic model based on the DCM distribution, which enables efficient retrieval and accurate ranking. Because the DCM distribution captures the dependency of repetitive word occurrences, the new probabilistic model based on this distribution is able to model the concavity of the score function more effectively. To avoid the empirical tuning of retrieval parameters, we design several parameter estimation algorithms to automatically set model parameters. Additionally, we propose a pseudo-relevance feedback algorithm based on the mixture modeling of the Dirichlet compound multinomial distribution to further improve retrieval accuracy. Finally, our experiments show that both the baseline probabilistic retrieval algorithm based on the DCM distribution and the corresponding pseudo-relevance feedback algorithm outperform the existing language modeling systems on several TREC retrieval tasks. The main objective of this research is to develop an effective probabilistic model based on the DCM distribution. A secondary objective is to provide a thorough understanding of the probabilistic retrieval model by a theoretical understanding of various text distribution assumptions.     

A prototype of an intelligent system for information retrieval: IOTA

Recent results in artificial intelligence research are of prime interest in various fields of computer science; in particular we think information retrieval may benefit from significant advances in this approach. Expert systems seem to be valuable tools for components of information retrieval systems related to semantic inference. The query component is the one we consider in this paper. IOTA is the name of the resulting prototype presented here, which is our first step toward what we call an intelligent system for information retrieval.After explaining what we mean by this concept and presenting current studies in the field, the presentation of IOTA begins with the architecture problem, that is, how to put together a declarative component, such as an expert system, and a procedural component, such as an information retrieval system. Then we detail our proposed solution, which is based on a procedural expert system acting as the general scheduler of the entire query processing. The main steps of natural language query processing are then described according to the order in which they are processed, from the initial parsing of the query to the evaluation of the answer. The distinction between expert tasks and nonexpert tasks is emphasized. The paper ends with experimental results obtained from a technical corpus, and a conclusion about current and future developments.     

Document management systems from current capabilities towards intelligent information retrieval: an overview

Documents circulating in paper form are increasingly being substituted by its electronic equivalent in the modern office today so that any stored document can be retrieved whenever needed later on. The office worker is already burdened with information overload, so effective and efficient retrieval facilities become an important factor affecting worker productivity. This paper first reviews the features of current document management systems with varying facilities to manage, store and retrieve either reference to documents or whole documents. Information retrieval databases, groupware products and workflow management systems are presented as developments to handle different needs, together with the underlying concepts of knowledge management. The two problems of worker finiteness and worker ignorance remain outstanding, as they are only partially addressed by the above-mentioned systems. The solution lies in a shift away from pull technology where the user has to actively initiate the request for information towards push technology, where available information is automatically delivered without user intervention. Intelligent information retrieval agents are presented as a solution together with a marketing scenario of how they can be introduced.     

Mining and modeling linkage information from citation context for improving biomedical literature retrieval

Mining linkage information from the citation graph has been shown to be effective in identifying important literatures. However, the question of how to utilize linkage information from the citation graph to facilitate literature retrieval still remains largely unanswered. In this paper, given the context of biomedical literature retrieval, we first conduct a case study in order to find out whether applying PageRank and HITS algorithms directly to the citation graph is the best way of utilizing citation linkage information for improving biomedical literature retrieval. Second, we propose a probabilistic combination framework for integrating citation information into the content-based information retrieval weighting model. Based on the observations of the case study, we present two strategies for modeling the linkage information contained in the citation graph. The proposed framework provides a theoretical support for the combination of content and linkage information. Under this framework, exhaustive parameter tuning can be avoided. Extensive experiments on three TREC Genomics collections demonstrate the advantages and effectiveness of our proposed methods.     

Intelligent information retrieval: diagnosing information need. : Part I. The theoretical framework for developing an intelligent IR tool

The research examines the notion that the principles underlying the procedure used by doctors to diagnose a patient's disease are useful in the design of “intelligent” IR systems because the task of the doctor is conceptually similar to the computer (or human) intermediary's task in “intelligent information retrieval”: to draw out, through interaction with the IR system, the user's query/information need. The research is reported in two parts. In Part II, an information retrieval tool is described which is based on “intelligent information retrieval” assumptions about the information user. In Part I, presented here, the theoretical framework for the tool is set out. This framework is borrowed from the diagnostic procedure currently used in medicine, called “differential diagnosis”. Because of the severe consequences that attend misdiagnosis, the operating principle in differential diagnosis is (1) to expand the uncertainty in the diagnosis situation so that all possible hypotheses and evidence are considered, then (2) to contract the uncertainty in a step by step fashion (from an examination of the patient's symptoms, through the patient's history and a physical (signs), to laboratory tests). The IR theories of Taylor, Kuhlthau and Belkin are used to demonstrate that these medical diagnosis procedures are already present in IR and that it is a viable model with which to design “intelligent” IR tools and systems.     

Relation of resemblance in information retrieval

A method using the amount of semantic information of query terms as weight in a fuzzy relation of resemblance is presented. The relation can be used to partially order documents in decreasing order of resemblance with the query. Large operational bibliographic data bases are used to test the validity of the approach.     

Artificial intelligence in information retrieval systems

A survey is given of the potential role of artificial intelligence in retrieval systems. Papers by Bush and Turing are used to introduce early ideas in the two fields and definitions for artificial intelligence and information retrieval for the purposes of this paper are given. A simple model of an information retrieval system provides a framework for subsequent discussion of artificial intelligence concepts and their applicability in information retrieval. Concepts surveyed include pattern recognition, representation, problem solving and planning, heuristics, and learning. The paper concludes with an outline of areas for further research on artificial intelligence in information retrieval systems.     

Knowledge-sparse and knowledge-rich learning in information retrieval

This paper reviews some aspects of the relationship between the large and growing fields of machine learning (ML) and information retrieval (IR). Learning programs are described along several dimensions. One dimension refers to the degree of dependence of an ML + IR program on users, thesauri, or documents. This paper emphasizes the role of the thesaurus in ML + IR work. ML + IR programs are also classified in a dimension that extends from knowledge-sparse learning at one end to knowledge-rich learning at the other. Knowledge-sparse learning depends largely on user yes-no feedback or on word frequencies across documents to guide adjustments in the IR system. Knowledge-rich learning depends on more complex sources of feedback, such as the structure within a document or thesaurus, to direct changes in the knowledge bases on which an intelligent IR system depends. New advances in computer hardware make the knowledge-sparse learning programs that depend on word occurrences in documents more practical. Advances in artificial intelligence bode well for knowledge-rich learning.     

Verbosity normalized pseudo-relevance feedback in information retrieval

Document length normalization is one of the fundamental components in a retrieval model because term frequencies can readily be increased in long documents. The key hypotheses in literature regarding document length normalization are the verbosity and scope hypotheses, which imply that document length normalization should consider the distinguishing effects of verbosity and scope on term frequencies. In this article, we extend these hypotheses in a pseudo-relevance feedback setting by assuming the verbosity hypothesis on the feedback query model, which states that the verbosity of an expanded query should not be high. Furthermore, we postulate the following two effects of document verbosity on a feedback query model that easily and typically holds in modern pseudo-relevance feedback methods: 1) the verbosity-preserving effect: the query verbosity of a feedback query model is determined by feedback document verbosities; 2) the verbosity-sensitive effect: highly verbose documents more significantly and unfairly affect the resulting query model than normal documents do. By considering these effects, we propose verbosity normalized pseudo-relevance feedback, which is straightforwardly obtained by replacing original term frequencies with their verbosity-normalized term frequencies in the pseudo-relevance feedback method. The results of the experiments performed on three standard TREC collections show that the proposed verbosity normalized pseudo-relevance feedback consistently provides statistically significant improvements over conventional methods, under the settings of the relevance model and latent concept expansion.     

Phrase structure rewrite systems in information retrieval

Operational level automatic indexing requires an efficient means of normalizing natural language phrases. Subject switching requires an efficient means of translating one set of authorized terms to another. A phrase structure rewrite system called a Lexical Dictionary is explained that performs these functions. Background, operational use, other applications and ongoing research are explained.     

Term relevance weights in on-line information retrieval

Considerable evidence exists to show that the use of term relevance weights is beneficial in interactive information retrieval. Various term weighting systems are reviewed. An experiment is then described in which information retrieval users are asked to rank query terms in decreasing order of presumed importance prior to actual search and retrieval. The experimental design is examined, and various relevance ranking systems are evaluated, including fully automatic systems based on inverse document frequency parameters, human rankings performed by the user population, and combinations of the two.     

Automatic query formulations in information retrieval.

Modern information retrieval systems are designed to supply relevant information in response to requests received from the user population. In most retrieval environments the search requests consist of keywords, or index terms, interrelated by appropriate Boolean operators. Since it is difficult for untrained users to generate effective Boolean search requests, trained search intermediaries are normally used to translate original statements of user need into useful Boolean search formulations. Methods are introduced in this study which reduce the role of the search intermediaries by making it possible to generate Boolean search formulations completely automatically from natural language statements provided by the system patrons. Frequency considerations are used automatically to generate appropriate term combinations as well as Boolean connectives relating the terms. Methods are covered to produce automatic query formulations both in a standard Boolean logic system, as well as in an extended Boolean system in which the strict interpretation of the connectives is relaxed. Experimental results are supplied to evaluate the effectiveness of the automatic query formulation process, and methods are described for applying the automatic query formulation process in practice.     

The weighted Condorcet fusion in information retrieval

The Condorcet fusion is a distinctive fusion method and was found useful in information retrieval. Two basic requirements for the Condorcet fusion to improve retrieval effectiveness are: (1) all component systems involved should be more or less equally effective; and (2) each information retrieval system should be developed independently and thus each component result is more or less equally different from the others. These two requirements may not be satisfied in many cases, then weighted Condorcet becomes a good option. However, how to assign weights for the weighted Condorcet has not been investigated.     

Using information gain to improve multi-modal information retrieval systems

Nowadays, access to information requires managing multimedia databases effectively, and so, multi-modal retrieval techniques (particularly images retrieval) have become an active research direction. In the past few years, a lot of content-based image retrieval (CBIR) systems have been developed. However, despite the progress achieved in the CBIR, the retrieval accuracy of current systems is still limited and often worse than only textual information retrieval systems. In this paper, we propose to combine content-based and text-based approaches to multi-modal retrieval in order to achieve better results and overcome the lacks of these techniques when they are taken separately. For this purpose, we use a medical collection that includes both images and non-structured text. We retrieve images from a CBIR system and textual information through a traditional information retrieval system. Then, we combine the results obtained from both systems in order to improve the final performance. Furthermore, we use the information gain (IG) measure to reduce and improve the textual information included in multi-modal information retrieval systems. We have carried out several experiments that combine this reduction technique with a visual and textual information merger. The results obtained are highly promising and show the profit obtained when textual information is managed to improve conventional multi-modal systems.     

信息系统实施用户抵制研究综述

综述IS领域解释用户抵制行为的研究,根据其研究对象,分为个体层抵制行为研究和组织层抵制行为研究,探讨其研究视角和理论发展,试图厘清IS领域中用户抵制研究的发展脉络,希望给出研究理论的全貌,为今后更好的进行用户抵制研究打下基础。     

Parsimonious translation models for information retrieval

In the KL divergence framework, the extended language modeling approach has a critical problem of estimating a query model, which is the probabilistic model that encodes the user’s information need. For query expansion in initial retrieval, the translation model had been proposed to involve term co-occurrence statistics. However, the translation model was difficult to apply, because the term co-occurrence statistics must be constructed in the offline time. Especially in a large collection, constructing such a large matrix of term co-occurrences statistics prohibitively increases time and space complexity. In addition, reliable retrieval performance cannot be guaranteed because the translation model may comprise noisy non-topical terms in documents. To resolve these problems, this paper investigates an effective method to construct co-occurrence statistics and eliminate noisy terms by employing a parsimonious translation model. The parsimonious translation model is a compact version of a translation model that can reduce the number of terms containing non-zero probabilities by eliminating non-topical terms in documents. Through experimentation on seven different test collections, we show that the query model estimated from the parsimonious translation model significantly outperforms not only the baseline language modeling, but also the non-parsimonious models.     

Ontology refinement for improved information retrieval

Ontologies are frequently used in information retrieval being their main applications the expansion of queries, semantic indexing of documents and the organization of search results. Ontologies provide lexical items, allow conceptual normalization and provide different types of relations. However, the optimization of an ontology to perform information retrieval tasks is still unclear. In this paper, we use an ontology query model to analyze the usefulness of ontologies in effectively performing document searches. Moreover, we propose an algorithm to refine ontologies for information retrieval tasks with preliminary positive results.     

Query-level loss functions for information retrieval

Many machine learning technologies such as support vector machines, boosting, and neural networks have been applied to the ranking problem in information retrieval. However, since originally the methods were not developed for this task, their loss functions do not directly link to the criteria used in the evaluation of ranking. Specifically, the loss functions are defined on the level of documents or document pairs, in contrast to the fact that the evaluation criteria are defined on the level of queries. Therefore, minimizing the loss functions does not necessarily imply enhancing ranking performances. To solve this problem, we propose using query-level loss functions in learning of ranking functions. We discuss the basic properties that a query-level loss function should have and propose a query-level loss function based on the cosine similarity between a ranking list and the corresponding ground truth. We further design a coordinate descent algorithm, referred to as RankCosine, which utilizes the proposed loss function to create a generalized additive ranking model. We also discuss whether the loss functions of existing ranking algorithms can be extended to query-level. Experimental results on the datasets of TREC web track, OHSUMED, and a commercial web search engine show that with the use of the proposed query-level loss function we can significantly improve ranking accuracies. Furthermore, we found that it is difficult to extend the document-level loss functions to query-level loss functions.