Business change is often accompanied by loss of continuity of experience. This has serious implications for the adaptation of an organisation's software since people with detailed knowledge of either the software or business processes may be unavailable to inform its adaptation. In many cases organisational memory will persist principally in the form of documents such as requirements specifications, operating procedures, regulatory standards, etc. These offer an important resource for informing what features of the software are redundant, need to be retained or can be reused. Exploiting this resource poses formidable problems, however, since it is often incomplete, poorly structured, poorly maintained and voluminous. This paper proposes that tools exploiting probabilistic natural language processing techniques offer the potential to ease these problems. Such tools are available, mature and have been proven in other domains. 1. Introduction Many organisations react to changes to their...

The introduction of computerized post-coordination has solved many of the problems of pre-coordinated subject access. However, the adoption of computerized post-coordination results in the loss of some pre-coordination benefits. Specifically, the effect of hiding terms within the context of others is lost in post-coordination which gives lead status to every document term, This results in spurious matches of terms out of context. Library patrons and Internet searchers are increasingly dissatisfied with subject access performance, in part because of unmanageably large retrieval sets. The need to enhance precision and limit the size of retrieval sets motivates this work which proposes partial coordination, an approach which incorporates the advantages of computer search with the ability of pre-coordination to limit spurious partial matches and thereby enhance precision.

. Exploratory data analysis seems to be a good tool for the acquisition and representation of the inherent knowledge in legal texts. The main difficulty besides the necessary input is the analysis of the various text and document structures. In our prototype CONCAT we use neural network technology to learn about the relations within the concept and document space of an existing domain. The results are quite encouraging because with existing input data a usable representation of the knowledge space can be obtained. 1 Introduction Exploratory data analysis seems to be a good tool for the representation of the inherent knowledge in legal texts. Existing legal information retrieval systems do not satisfy the demands of lawyers because they provide only a syntactic representation of the legal data (e.g. statutes, treaties, court decisions or literature). Advanced formalisations of legal knowledge exist in the form of legal expert systems or conceptual information retrieval systems. The ma...

. This paper is concerned with the application of Kohonen's self-organising feature map to legal knowledge acquisition. More precisely, the map is used for analysis of legal thesauri which are obtained by means of connotation analysis of the individual document descriptors. The outcome of the learning process of the artificial neural network is further used to distinguish between precise legal terms and terms with rather fuzzy meaning. INTRODUCTION The formalisation of legal data, e.g. statutes, court decisions or treaties, constitutes a necessary prerequisite for advanced legal information processing. Common approaches to cope with this problem are first, rewriting the law as a logic program and second, performing linguistic analysis of the legal language. The former approach represents a harsh simplification of the complexity of legal systems and thus leads to severe limitations as for example the open texture problem. Inherent to the latter approach is the obvious fact that so far...

Only by continuous self-appraisal can a large information system make itself responsive to the needs of the scientific community. Concluding sentence in Lancaster (1969) The main objective of information retrieval (IR) systems is to retrieve information or information objects relevant to user requests and possible needs. In IR tests, retrieval effectiveness is established by comparing IR systems retrievals (systems relevance) with users’ or user surrogates ’ assessments (user relevance), where user relevance is treated as the gold standard for performance evaluation. Relevance is a human notion, and establishing relevance by humans is fraught with a number of problems—inconsistency in judgment being one of them. The aim of this critical review is to explore the relationship between relevance on the one hand and testing of IR systems and procedures on the other. Critics of IR tests raised the issue of validity of the IR tests because they were based on relevance judgments that are inconsistent. This review traces and synthesizes experimental studies dealing with (1) inconsistency of relevance judgments by people, (2) effects of such inconsistency on results of IR tests and (3) reasons for retrieval failures. A historical context for these studies and for IR testing is provided including an assessment of Lancaster’s (1969) evaluation of MEDLARS and its unique place in the history of IR evaluation.

. Traditional information retrieval systems do not satisfy the lawyers' demands because they provide only syntactic representation of legal data. The bottleneck for the creation of the more promising conceptual information retrieval systems is the time-consuming knowledge acquisition. The best solution is the representation of legal knowledge by simple linguistic tools, statistics and neural networks. In our prototype KONTERM we represent legal knowledge about concepts and documents by a knowledge base which is structured by statistical and connectionist methods. In future, this knowledge base will be used to filter legal knowledge from documents. to provide a solution for this problem where the emphasis is on the representation of relevant information in the form of text patterns. Such text patterns can be used to formalise the expert knowledge about legal language. Lawyers have formed definite concepts of human beings, objects, and processes by use of methods of abstraction and log...

Knowledge acquisition constitutes the bottleneck for the creation of legal expert systems. Legal language must be formalised to such a degree that it can be processed automatically. We deal with this problem by supporting the process of creating a selective thesaurus for a legal information system which can be seen as prerequisite for further knowledge processing. This selectivity is obtained by means of connotation analysis of the individual descriptors which makes it possible to detect hidden word meanings and to distinguish between precise legal terms and words with fuzzy meaning. Within the prototype system CONCAT we applied both a statistical and a connectionist approach to connotation analysis and performed a comparative evaluation of the achieved results. 1 Introduction Advanced use of information technology in the legal field requires formalisation of the legal data (e.g. statutes, treaties, court decisions or literature). Two main approaches are concerned with this ...

. Besides working with large amount of documents, information retrieval has to deal with the uncertainty confronting all natural languages: homonymy, synonymy and polysemy; majors hurdles that every automatic natural language processing must deal with. In order to encourage the discovery of better solutions to these problems and to encourage improved understanding of the matching process between query and documents, our approach is to view retrieval mechanisms as an inference process involving uncertainty. In this context, a fundamental question is the choice of an adequate framework which will allow us to: (1) represent the various types of uncertain knowledge; (2) combine various sources of evidence about query or document content; and (3) come up with efficient and sound techniques for making the needed inferences. To meet these criteria, we suggest using probabilistic argumentation systems (PAS) which marry propositional logic with probability theory such that we can deal with unce...

We present an approach to increasing the effectiveness of ranked-output retrieval systems that relies on graphical display and user manipulation of "views" of retrieval results, where a view is the subset of retrieved documents that contain a specified subset of query terms. This approach has been implemented in a system named VIEWER (VIEwing WEb Results), acting as an interface to available search engines. An experimental evaluation of the performance of VIEWER in contrast to AltaVista is the major focus of the paper. We first report the results of an experiment on single, short query searches where VIEWER, used as an interactive ranking system, markedly outperformed AltaVista. We then concentrate on a more realistic searching scenario, involving free query formulation, unconstrained selection of retrieval results, and possibility of query reformulation. We report the results of an experiment where the use of VIEWER, compared to AltaVista, seemed to shift the user effort from inspection to evaluation of results, increasing retrieval effectiveness and user satisfaction. In particular, we found that the VIEWER users retrieved half as many nonrelevant documents as the AltaVista users while retrieving a comparable number of relevant documents.

We can distinguish two types of web search engines: general use ones that index and search all the web, and site-specific ones that are provided by individual websites for local searching. A comparisson of the effectiveness of the two types allows search engine users to choose the right engine and organizations to decide whether they should develop their own search software or purchase the search function as a service. We evaluate the performance of two general purpose search engines and 10 sitespecific ones. The criteria we used are precision and relative recall. We entered 20 queries in each website’s search engine and evaluated the first 10 links. According to the results, Google is in most cases the most efficient search engine. However, in some cases general purpose search engines do not index the website’s content as well as a site-specific engine. 1.