Models of document indexing and document retrieval have been extensively studied. The integration of these two classes of models has been the goal of several researchers but it is a very difficult problem. We argue that much of the reason for this is the lack of an adequate indexing model. This suggests that perhaps a better indexing model would help solve the problem. However, we feel that making unwarranted parametric assumptions will not lead to better retrieval performance. Furthermore, making prior assumptions about the similarity of documents is not warranted either. Instead, we propose an approach to retrieval based on probabilistic language modeling. We estimate models for each document individually. Our approach to modeling is non-parametric and integrates document indexing and document retrieval into a single model. One advantage of our approach is that collection statistics which are used heuristically in many other retrieval models are an integral part of our model. We have...

The 2–Poisson model for term frequencies is used to suggest ways of incorporating certain variables in probabilistic models for information retrieval. The variables concerned are within-document term frequency, document length, and within-query term frequency. Simple weighting functions are developed, and tested on the TREC test collection. Considerable performance improvements (over simple inverse collection frequency weighting) are demonstrated. 1

The naive Bayes classifier, currently experiencing a renaissance in machine learning, has long been a core technique in information retrieval. We review some of the variations of naive Bayes models used for text retrieval and classification, focusing on the distributional assump- tions made about word occurrences in documents.

The paper combines a comprehensive account of the probabilistic model of retrieval with new systematic experiments on TREC Programme material. It presents the model from its foundations through its logical development to cover more aspects of retrieval data and a wider range of system functions. Each step in the argument is matched by comparative retrieval tests, to provide a single coherent account of a major line of research. The experiments demonstrate, for a large test collection, that the probabilistic model is effective and robust, and that it responds appropriately, with major improvements in performance, to key features of retrieval situations.

We introduce and create a framework for deriving probabilistic models of Information Retrieval. The models are nonparametric models of IR obtained in the language model approach. We derive term-weighting models by measuring the divergence of the actual term distribution from that obtained under a random process. Among the random processes we study the binomial distribution and Bose–Einstein statistics. We define two types of term frequency normalization for tuning term weights in the document–query matching process. The first normalization assumes that documents have the same length and measures the information gain with the observed term once it has been accepted as a good descriptor of the observed document. The second normalization is related to the document length and to other statistics. These two normalization methods are applied to the basic models in succession to obtain weighting formulae. Results show that our framework produces different nonparametric models forming baseline alternatives to the standard tf-idf model.

In this paper, an introduction and survey over probabilistic information retrieval (IR) is given. First, the basic concepts of this approach are described: the probability ranking principle shows that optimum retrieval quality can be achieved under certain assumptions; a conceptual model for IR along with the corresponding event space clarify the interpretation of the probabilistic parameters involved. For the estimation of these parameters, three different learning strategies are distinguished, namely query-related, document-related and description-related learning. As a representative for each of these strategies, a specific model is described. A new approach regards IR as uncertain inference; here, imaging is used as a new technique for estimating the probabilistic parameters, and probabilistic inference networks support more complex forms of inference. Finally, the more general problems of parameter estimation, query expansion and the development of models for advanced document representations are discussed.

Abstract- in this article three retrieval models for probabilistic indexing are described along with evaluation results for each. First is the binary independence indexing @II) model, which is a generalized version of the Maron and Kuhns indexing model. In this model, the indexing weight of a descriptor in a document is an estimate of the proba-bility of relevance of this document with respect to queries using this descriptor. Sec-ond is the retrieval-with-probabilistic-indexing (RPI) model, which is suited to different kinds of probabilistic indexing. For that we assume that each indexing scheme has its own concept of “correctness ” to which the probabilities relate. In addition to the prob-abilistic indexing weights, the RPI model provides the possibility of reIevance weight-ing of search terms. A third mode1 that is similar was proposed by Croft some years ago as an extension of the binary independence retrieval model but it can be shown that this model is not based on the probabilistic ranking principle. The probabilistic indexing weights required for any of these models can be provided by an application of the Darm-stadt indexing approach (DIA) for indexing with descriptors from a controlled vocabu-Iary. The experimental results show signi~cant improvements over retrieval with binary indexing. Finally, suggestions are made regarding how the DIA can be applied to prob-abilistic indexing with free text terms. 1.

In this paper the score distributions of a number of text search engines are modeled. It is shown empirically that the score distributions on a per query basis may be fitted using an exponential distribution for the set of non-relevant documents and a normal distribution for the set of relevant documents. Experiments show that this model fits TREC-3 and TREC-4 data for not only probabilistic search engines like INQUERY but also vector space search engines like SMART for English. We have also used this model to fit the output of other search engines like LSI search engines and search engines indexing other languages like Chinese. It is then shown that given a query for which relevance information is not available, a mixture model consisting of an exponential and a normal distribution can be fitted to the score distribution. These distributions can be used to map the scores of a search engine to probabilities. We also discuss how the shape of the score distributions arise given certain assumptions about word distributions in documents. We hypothesize that all 'good' text search engines operating on any language have similar characteristics. This model has many possible applications. For example, the outputs of different search engines can be combined by averaging the probabilities (optimal if the search engines are independent) or by using the probabilities to select the best engine for each query. Results show that the technique performs as well as the best current combination techniques. This material is based on work supported in part by the National Science Foundation, Library of Congress and Department of Commerce under cooperative agreement number EEC-9209623, in part by the National Science Foundation under grant numbers IRI-9619117 and IIS-9909073, in part by N...

Shannon (1948) showed that a wide range of practical problems can be reduced to the problem of estimating probability distributions of words and ngrams in text. It has become standard practice in text compression, speech recognition, information retrieval and many other applications of Shannon's theory to introduce a "bag-of-words" assumption. But obviously, word rates vary from genre to genre, author to author, topic to topic, document to document, section to section, and paragraph to paragraph. The proposed Poisson mixture captures much of this heterogeneous structure by allowing the Poisson parameter theta to vary over documents subject to a density function phi. phi is intended to capture dependencies on hidden variables such [as] genre, author, topic, etc. (The Negative Binomial is a well-known special case where phi is a Gamma distribution.) Poisson mixtures fit the data better than standard Poissons, producing more accurate estimates of the variance over documents (sigma^2), entropy (H), inverse document frequency (IDF), and adaptation (Pr(x>=2|x>=1)).

Discriminative models have been preferred over generative models in many machine learning problems in the recent past owing to some of their attractive theoretical properties. In this paper, we explore the applicability of discriminative classifiers for IR. We have compared the performance of two popular discriminative models, namely the maximum entropy model and support vector machines with that of language modeling, the state-of-the-art generative model for IR. Our experiments on ad-hoc retrieval indicate that although maximum entropy is significantly worse than language models, support vector machines are on par with language models. We argue that the main reason to prefer SVMs over language models is their ability to learn arbitrary features automatically as demonstrated by our experiments on the home-page finding task of TREC-10.