The profile-query relationship
β Scribed by Shepherd, Michael A. ;Phillips, W. J.
- Book ID
- 102652855
- Publisher
- John Wiley and Sons
- Year
- 1986
- Tongue
- English
- Weight
- 693 KB
- Volume
- 37
- Category
- Article
- ISSN
- 0002-8231
No coin nor oath required. For personal study only.
β¦ Synopsis
This research defines the relationship between a user profile and a user query in terms of the relationship between the clusters of documents that are retrieved by the profile and by the query, respectively. The document clusters were generated by performing retrieval at various threshold levels on profiles and related queries. Profiles were artificially generated based on documents by an author. Queries were artificially generated to cover the full range of the relationship as defined by the cosine similarity measure. The resulting clusters were examined and it was found that, given a desired retrieval threshold, cluster similarity and cluster overlap can be expressed as linear functions of the similarity existing between the original pairs of profiles and queries.
π SIMILAR VOLUMES
More and more relational database systems for microcomputers are providing visual interfaces, such as query by example (QBE). At the same time, experimental studies are providing more support for the theory that the end users can perform better with the entity relationship (ER) model than with the r
In experimental studies on query languages, subjects are required to write queries using di!erent query languages. User query performance is usually measured by query accuracy. There is no clearly de"ned objective method of applying "ndings to other queries. This study examines the suitability of us