Product Recommendation System by Approximate Search Based on Manhattan Distance Measurement

ผู้วิจัย

VASUPON PHUEAKNUMPOL, ZAGON BUDSABONG, KITTISAK KERDPRASOP, NITTAYA KERDPRASOP

บทคัดย่อ

Database search is an important technology to assist users finding specific information from the vast amount of contents. However, the method of searching is limited in a way that the result must be a hundred percent matching the needs of users. Some attributes that may be required but are missing from the conditions of users’ query make the partial relevant data instances be eliminated. This situation occurs when users input an incorrect condition in searching or input a rough predicate. The consequence is that users will not find any result or find instead a not-required result. From this point of view, we propose an approximate search method to find instances from the database. Approximation approach is to compare the distance between the points of input data and the data that have the potential to be a desired output. The comparison is based on the Manhattan-distance computation method. The results of each query search can be varied depending on the user’s condition.

บรรณานุกรม

[1] S. Sonnum, S. Thaithieng, S. Ano, K. Kusolchu, and N. Kerdprasop, Approximate Web Database Search Based on Euclidean Distance Measurement, Proc. IMECS 2011, March 2011. [2] S. Pramanik and J. Li, Fast Approximate Search Algorithm for Nearest Neighbor Queries in High Dimensions, IEEE Computer Society, 1999. [3] S. Wu, H.T. Chang, T.C. Hsu, and P.S. Liu, Approximate Keyword Search in Web Search Engines, Digital Information Management, December 2006 [4] Y. Oike, M. Ikeda, and K. Asadat, A Word-Parallel Digital Associative Engine with Wide Search Range Based on Manhattan Distance, IEEE 2004 Custom Integrated Circuits Conference, 2004.