Browsing in a digital library collecting linearly arranged documents

Abstract

A method of assisting a user in finding the required documents effectively is proposed. A user being informed which documents are worth examining can browse in a digital library (DL) in a linear fashion. Computational evaluations were carried out, and a DL and its navigator are designed and constructed.

Handling document expressions for capturing conceptual query–document relations

“…In order to enhance the general use of the formulated methods, they must be applied to various collections other than the ones used in the experiments and their effectiveness ascertained. In addition, there are also plans to incorporate the methods into processes [23,24] that feed back the results of matching judgments to learn search objectives of users.…”

“…Figure 5 shows the results. When there is a large-scale conceptual hierarchy, a mechanism [23,24] can be utilized that allows the identifiers to be visualized in a linear arrangement to only emphasize and display locations that should be searched.…”

“…Search results according to initial queries are accumulated in electronic libraries for information searches which have spread throughout the Internet and users can find what they are searching for while browsing through these libraries [21,30]. In this environment, if the contents of a document judged to match in a browsing process are utilized for learning of a search that was unclear from the start, truly desired information can be efficiently found [23,24]. For this case, however, the manner in which the user rapidly judges matches in many documents is the key.…”

Browsing‐based conceptual information retrieval incorporating dictionary term relations, keyword association, and a user's interest

“…The retrieval performance can be improved by considering further a user's interest and the keyword association with the documents. A learning module does this job, during the process of browsing the arranged documents in a pDL, by learning interactively and incrementally a user's interest in the form of a local dictionary (Qu, Sato, Nakashima, & Ito, 2001). In a local dictionary the documents already judged to be relevant or non‐relevant are stored in the separate sets.…”

“…To capture the keyword association and a user's interest we here organize, by using a simple interactive concept‐learning algorithm (Aha, Kibler, & Albert, 1991), a local dictionary (named in contrast to a global dictionary) that incrementally collects two sets of the judged relevant and non‐relevant documents (Qu, Sato, Nakashima, & Ito, 2001). By the information in a local dictionary a user is informed “which documents are worth examining in more detail” when he/she browses in a pDL.…”