On More Efficient Searching Methods for a Corporate Knowledge Database

摘要

This paper discusses possible assisting approaches with that the search engine indicates additional keywords for searching the desired documents after the first keyword is specified by the user. Effectiveness of a new idea to utilize a frequency distribution of terminologies used in all the database documents was evaluated, showing a promising indication that the approach is useful for suggesting good candidate terminologies for secondary keywords. The paper indicates necessary steps towards building of a more efficient search engine including establishment of a solid mathematical or statistical foundation for determining a group of meaningful terminologies from any arbitrary database, and combination with other approaches for augmenting searching performance.