English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Consider Figure 1. Which three documents are most similar to the query,Q? It depends, of course, on the similarity measure that ones use. However, when this figure has been shown informally at professional conferences, the majority of the audience—up to 90%—chooses documents D2, D4, andD6, a choice obviously based on distance rather than angle. Why, then, is the cosine measure so dominant in vector retrieval research? The cosine measure and a distance measure are clearly sensitive to different aspects of documents. Yet a search of the literature on vector retrieval has not revealed a clear argument for the use of the cosine measure rather than a distance measure. In this article we do not propose to make this argument, but rather to present a user interface that incorporates an angular and distance measure, thus allowing the user to judge documents by both measures. Information retrieval has developed rapidly in recent years. Development is reflected in following aspects: the types and number of information systems have increased dramatically, the ways of information organization have become varied and complicated, and information retrieval evaluation models and means have become not only more powerful but also more sophisticated. However, users hope that the systems are more flexible; they do not care about the complicated mathematical foundations or the calculation of model metrics. Under the circumstances, some of main trends in the information retrieval field are: (a) how to make systems more user friendly, more transparent to users; and (b) how to provide effective and convenient mechanisms for users to understand the information retrieval process, compare effectiveness of the different evaluation models, dynamically adjust search strategy, and finally achieve satisfactory retrieval results. To visualize the information retrieval process, display the document distribution in a certain scope, and offer the means to browse the relevant documents are some of the effective ways to achieve the aims. Now, many researchers on the visualization of information retrieval have achieved significant results (Ahlberg & Shneiderman, 1994; Arent et al., 1984; Chalmers & Chitson, 1992; Furnas & Rauch, 1998; Hearst & Karadi, 1997; Heath et al., 1995; Allen, 1998; Benford et al., 1995; Bradley, 1996; Brodbeck et al., 1997; Chalmers, 1993; Hearst, 1995; Helfman, 1994; Koshman, 1996; Nuchprayyoon, 1996). Here, we present a tool for visualization of information retrieval. It provides an interactive visual interface for the users to filter irrelevant documents, view retrieval results, and control the retrieval process on the basis of the distance between document and query as well as a particular angle in document vector space. So far, there is no visualization model that can simultaneously reflect the two basic measures, distance and angle, which present two distinct methods to judge similarity. DARE is the first visualization tool that can provide such functions. Like other models, theDARE model has its inherent weaknesses, ambiguity being one of them. Ambiguous interpretation in the visualization space is a side-effect of the reduction of dimensionality of the document vector space.

DARE: Distance and Angle Retrieval Environment: A Tale of the Two Measures