การคัดกรองเอกสารที่สืบคนโดยการแปลงน้ําหนัก-ระยะหาง Filtering Search Document using Weight-Distance Transformation

Searching information from the Internet via available search engines is often overwhelmed by myriad of resulting documents that are mostly irrelevant. The problem lies in the use of proper keywords arranged in the right order. This paper proposes an effective filtering approach that exploits various existing techniques through a sequence of transformations. The proposed approach employs ontology tree as a basis for keyword matching, thereby intermediate search results can be stochastically classified and clustered into their respective domains. Subsequent transformations are applied to map the results on to a document space. A user-defined ideal document is utilized to establish measured weights obtained from the Euclidean distance between a given document and the ideal document. This process is repeated until a predetermined threshold is reached, giving rise to a straightforward document filtering technique. The direct benefit from the proposed approach is low number of search documents. This, in turn, entails continual research on reducing energy consumption as the volume of information to be processed and transmitted reduces. This is essential and conforming to today’s energy saving directives. It’s just a matter of technology transfer to realize the proposed approach in commercial applications.