Keith van Rijsbergen

Incorporating user search behavior into relevance feedback

In this paper, we present five user experiments on incorporating behavioral information into the relevance feedback process. In particular, we concentrate on ranking terms for query expansion and selecting new terms to add to the users query. Our experiments are an attempt to widen the evidence used for relevance feedback from simply the relevant documents to include information on how users are searching. We show that this information can lead to more successful relevance feedback techniques. We also show that the presentation of relevance feedback to the user is important in the success of relevance feedback.

What can quantum theory bring to information retrieval

The probabilistic formalism of quantum physics is said to provide a sound basis for building a principled information retrieval framework. Such a framework can be based on the notion of information need vector spaces where events, such as document relevance or observed user interactions, correspond to subspaces. As in quantum theory, a probability distribution over these subspaces is defined through weighted sets of state vectors (density operators), and used to represent the current view of the retrieval system on the user information need. Tensor spaces can be used to capture different aspects of information needs. Our evaluation shows that the framework can lead to acceptable performance in an ad-hoc retrieval task. Going beyond this, we discuss the potential of the framework for three active challenges in information retrieval, namely, interaction, novelty and diversity.

Combining and selecting characteristics of information use

Ruthven, Lalmas, and van Rijsbergen use traditional term importance measures like inverse document frequency, noise, based upon in-document frequency, and term frequency supplemented bytheme value which is calculated from differences of expected positions of words in a text from their actual positions, on the assumption that even distribution indicates term association with a main topic, andcontext,which is based on a query terms distance from the nearest other query term relative to the average expected distribution of all query terms in the document. They then define document characteristics likespecificity, the sum of all idf values in a document over the total terms in the document, or document complexity, measured by the documents averageidf value; and information to noise ratio, info-noise, tokens after stopping and stemming over tokens before these processes, measuring the ratio of useful and non-useful information in a document. Retrieval tests are then carried out using each characteristic, combinations of the characteristics, and relevance feedback to determine the correct combination of characteristics. A file ranks independently of query terms by both specificity and info-noise, but if presence of a query term is required unique rankings are generated.

Supporting polyrepresentation in a quantum-inspired geometrical retrieval framework

The relevance of a document has many facets, going beyond the usual topical one, which have to be considered to satisfy a users information need. Multiple representations of documents, like user-given reviews or the actual document content, can give evidence towards certain facets of relevance. In this respect polyrepresentation of documents, where such evidence is combined, is a crucial concept to estimate the relevance of a document. In this paper, we discuss how a geometrical retrieval framework inspired by quantum mechanics can be extended to support polyrepresentation. We show by example how different representations of a document can be modelled in a Hilbert space, similar to physical systems known from quantum mechanics. We further illustrate how these representations are combined by means of the tensor product to support polyrepresentation, and discuss the case that representations of documents are not independent from a user point of view. Besides giving a principled framework for polyrepresentation, the potential of this approach is to capture and formalise the complex interdependent relationships that the different representations can have between each other.

Quantum Mechanics and Information Retrieval

In this chapter, we describe how the intersection between information retrieval and the quantum mechanical framework has been implemented. We select and present in no predefined order the most significant contributions to the implementation of this intersection; some contributions appear to be less mature than others; however, we decided to include them since they are sources of future work. Other contributions might appear less “quantum inspired” than others; however, each research work contains concepts and tools that are somehow linked to the quantum mechanical framework illustrated in the book. In the end, the contributions reported in this chapter cover a wide range of issues, from modeling issues to user interaction issues. The chapter ends with suggestions of further reading.

A Logical Model of Information Retrieval Based on Situation Theory

We use Logics to model relevance in Information Retrieval: a document is relevant to a query if a formula q representing the query can be inferred from a formula d representing the document. Thus to infer is to retrieve, but because of the nature of aboutness often the inference is uncenain. Using a framework based on Situation Theory, the representation of documents and queries, inference, semantic and pragmatic aspects of information can be modelled formally.

Towards quantum-based DB+IR processing based on the principle of polyrepresentation

The cognitively motivated principle of polyrepresentation still lacks a theoretical foundation in IR. In this work, we discuss two competing polyrepresentation frameworks that are based on quantum theory. Both approaches support different aspects of polyrepresentation, where one is focused on the geometric properties of quantum theory while the other has a strong logical basis. We compare both approaches and outline how they can be combined to express further aspects of polyrepresentation.

Automatically generating queries for prior art search

This paper outlines our participation in CLEF-IPs 2009 prior art search task. In the tasks initial year our focus lay on the automatic generation of effective queries. To this aim we conducted a preliminary analysis of the distribution of terms common to topics and their relevant documents, with respect to term frequency and document frequency. Based on the results of this analysis we applied two methods to extract queries. Finally we tested the effectiveness of the generated queries on two state of the art retrieval models.

Semantic Information Retrieval

Semantic Information Theory (SIT) is concerned with studies in Logic and Philosophy on the use of the term information, “in the sense in which it is used of whatever it is that meaningful sentences and other comparable combinations of symbols convey to one who understands them” (Hintikka, 1970). Notwithstanding the large scope of this description, SIT has primarily to do with the question of how to weigh sentences according to their informative content. The main difference with conventional information theory is that information is not conveyed by an ordered sequence of binary symbols, but by means of a formal language in which logical statements are defined and explained by a semantics. The investigation of SIT concerns two research directions: the axiomatisation of the logical principles for assigning probabilities or similar weighting functions to logical sentences and the relationship between information content of a sentence and its probability.

Knowledge modeling in prior art search

This study explores the benefits of integrating knowledge representations in prior art patent retrieval. Key to the introduced approach is the utilization of human judgment available in the form of classifications assigned to patent documents. The paper first outlines in detail how a methodology for the extraction of knowledge from such an hierarchical classification system can be established. Further potential ways of integrating this knowledge with existing Information Retrieval paradigms in a scalable and flexible manner are investigated. Finally based on these integration strategies the effectiveness in terms of recall and precision is evaluated in the context of a prior art search task for European patents. As a result of this evaluation it can be established that in general the proposed knowledge expansion techniques are particularly beneficial to recall and, with respect to optimizing field retrieval settings, further result in significant precision gains.