Although a graph can be a useful device for visualizing complex relationships, drawings of large graphs can be difficult to comprehend. As one remedy, we formulated a novel generalized approach for generating fisheye views of nested graphs with multiple variable focal points, and devised an algorithm that creates fisheye views in the absence of application specific distance metrics. Previous solutions produced fisheye views by filtering or distorting drawings of graphs. Since these approaches relied on geometric notions of distance, they could only be applied effectively in limited cases. By contrast, our approach treats fisheye view generation as a phase that precedes graph layout, rather than as a technique that alters an existing drawing, and does not suffer these drawbacks. 1 Introduction The graph topovisual formalism [6] is well-suited to the task of visualizing complex relations on a set of elements [2, 13, 15, 17, 20]. Large information spaces, however, must be depicted by lar...

Current best-match ranking (BMR) systems perform well but cannot handle word mismatch between a query and a document. The best known alternative ranking method, hierarchical clustering-based ranking (HCR), seems to be more robust than BMR with respect to this problem, but it is hampered by theoretical and practical limitations. We present an approach to document ranking that explicitly addresses the word mismatch problem by exploiting interdocument similarity information in a novel way. Document ranking is seen as a querydocument transformation driven by a conceptual representation of the whole document collection, into which the query is merged. Our approach is based on the theory of concept (or Galois) lattices, which, we argue, provides a powerful, well-founded, and computationallytractable framework to model the space in which documents and query are represented and to compute such a transformation. We compared information retrieval using concept lattice-based ranking (CLR) to BMR and HCR. The results showed that HCR was outperformed by CLR as well as by BMR, and suggested that, of the two best methods, BMR achieved better performance than CLR on the whole document set while CLR compared more favorably when only the first retrieved documents were used for evaluation. We also evaluated the three methods' specific ability to rank documents that did not match the query, in which case the superiority of CLR over BMR and HCR (and that of HCR over BMR) was apparent.

This paper presents the results of experiments carried out with a set of 4,000 medical discharge summaries in which were recognised 1,962 attributes from the Unified Medical Language System (UMLS). In this domain, the objects are medical documents (4,000) and the attributes are UMLS terms extracted from the documents (1,962). When Formal Concept Analysis is used to iteratively analyse and visualize this data, complexity and scalability become critically important. Although the amount of data used in this experiment is small compared with the size of primary memory in modern computers, the results are still important since the probability distributions which determine the efficiencies are likely to remain stable as the size of the data is increased. Our work presents two outcomes, firstly we present a methodology for exploring knowledge in text documents using Formal Concept Analysis by employing conceptual scales created as the result of direct manipulation of a line diagram. The conceptual scales lead to small derived purified contexts that are represented using nested line diagrams. Secondly, we present an algorithm for the fast determination of purified contexts from a compressed representation of the large formal context. Our work draws on existing encoding and compression techniques to show how rudimentary data analysis can lead to substantial efficiency improvements to knowledge visualisation. c fl 1993 Blackwell Publishers, 238 Main Street, Cambridge, MA 02142, USA, and 108 Cowley Road, Oxford, OX4 1JF, UK. SCALABILITY IN FORMAL CONCEPT ANALYSIS 3

In this paper we present a comprehensive approach to automatic organization and hybrid navigation of text databases. An organizing stage first builds a particular lattice representation of the data, through text indexing followed by lattice clustering of the indexed texts. The lattice representation, then, supports the navigation stage of the system, a visual retrieval interface that combines three main retrieval strategies: browsing, querying, and bounding. Browsing and querying are used to search the retrieval space, bounding is used to restrict it based on the information that users have, or get during their interaction with the system. We show that such a hybrid paradigm permits high flexibility in trading off information exploration and retrieval and, in addition, has good retrieval performance. We compared information retrieval using lattice-based hybrid navigation with conventional Boolean querying. The results of an experiment conducted on two medium-sized bibliographic databases showed that the performance of lattice retrieval was comparable to or better than Boolean retrieval

Document retrieval is a highly interactive process dealing with large amounts of information. Visual representations can provide both a means for managing the complexity of large information structures and an interface style well suited to interactive manipulation. The system we have designed utilizes visually displayed graphic structures and a direct manipulation interface style to supply an integrated environment for retrieval. A common visually displayed network structure is used for query, document content, and term relations. A query can be modified through direct manipulation of its visual form by incorporating terms from any other information structure the system displays. An associative thesaurus of terms and an inter-document network provide information about a document collection that can complement other retrieval aids. Visualization of these large data structures makes use of fisheye views and overview diagrams to help overcome some of the inherent difficulties of orientation and navigation in large information structures.

this paper attempts to strike a middle road allowing the user to construct and modify scales in response to learning information about the data. It is novel in that it allows the user to de ne a hierarchy over the search terms and presents the user with a dynamic environment for the creation and modi cation of scales by the user.

This paper discusses a novel method called active browsing which increases the speed and accuracy with which a user may browse libraries for reusable software. Information inferred solely from the user’s normal actions is employed by the system to locate software items relevant to the user’s search goal. This paper describes our active browsing system and illustrates its operation with an example using typical browsing steps. An experiment, using an automated browsing agent, is described demonstrating that active browsing accelerates search. 1

The Two-Level Hypermedia Paradigm sees an Information Retrieval System as consisting of a document network (the Hyperbase) and a descriptor (term) network (the Hyperindex). Query by Navigation is a process whereby the searcher gives a description of the Information Need by travelling through the descriptor network. This paper presents a formalism for expressing the effects of traversing the Hyperindex on the elements of the Hyperindex. This formalism makes use of probabilities for modelling the searcher's behavious. The events which can occur during the search process are discussed and modelled. Some important properties, which are reasonable to demand of a retrieval system, can be proven to be valid if this formalism is adopted. A mechanism for assigning a measure of relevance to documents is presented. This uses the formalism mentioned above. An example will show the effectiveness of The aspect of relevance feedback and its role in Query by Navigation is introduced by examining the d...

Abstract. It can be difficult to automatically generate “nice ” graphical representations for concept lattices from lexical databases, such as Roget’s Thesaurus, because the data sources tend to be large and complex. This paper discusses a variety of “data weeding ” techniques that can be applied in order to reduce the size of a concept lattice, first in general, and then with respect to Roget’s Thesaurus. The aim is that resulting lattices should display neither too much, nor too little information, independently of which search terms have been entered by a user. 1

Abstract. The following paper proposes a new kind of relevance feedback. It shows how so called query profiles can be employed for disambiguation and clarification. Query profiles provide useful summarized previews on the retrieved answers to a given query. They outline ambiguity in the query and when combined with appropriate means of interactivity allow the user to easily adapt the final ranking. Statistical analysis of the profiles even enables the retrieval system to automatically suggest search restrictions or preferences. The paper shows a preliminary experimental study of the proposed feedback methods within the setting of TREC’s interactive HARD track. 1