Conference presentation
Structuralizing a natural language knowledge base: A case-based reasoning system for library reference services
Drexel University. College of Information Science and Technology. Research Day Posters.
03 Jul 2007
Abstract
Reference librarians are the connection between the library's knowledge repository and the public. Hummelshoj [1, p.13] argues that "development of community information services with maximum human support is crucial for users' access to and use of information". However, the tremendous growth in information resources presents challenges to the reference librarians who need to find ways to provide searching assistance to the public. We propose that a Case-based Reasoning (CBR) type of retrieval could be useful in helping the librarians find similar types of questions and the previously worked-out answers could be reused and adapted when necessary. The framework of CBR, an Artificial Intelligence methodology, is premised on the natural form of human reasoning and learning from prior experience [3]. The knowledge base consists of individual cases built from episodes encountered previously [4]. Traditional CBR has offered beneficial results when the environment is well structured with adequately defined processes or case characteristics; however this leaves out areas where the experiences reuse are very rich but available only in natural language text [2]. This paper reports on the results of a project using Textual CBR to organize and represent natural language knowledge base in such a way that it could be offered as suggested solutions. We tested the possibility of categorizing reference questions in a knowledge base into case-groups. Here, CBR is used in a novel domain where there is a high level of variability among the potentially high volume of cases and where the acquired knowledge is in the form of free text of undetermined length. Pairs of reference questions and answers (QAs) are categorized into types in the knowledge base, and the solutions are the paired answers of the retrieved questions.
Metrics
12 File views/ downloads
6 Record Views
Details
- Title
- Structuralizing a natural language knowledge base: A case-based reasoning system for library reference services
- Creators
- Yen Bui (Author) - Drexel University (1970-)
- Publication Details
- Drexel University. College of Information Science and Technology. Research Day Posters.
- Resource Type
- Conference presentation
- Language
- English
- Academic Unit
- DU; College of Information Science and Technology (1995-2013)
- Identifiers
- 991014632083304721