We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
HYBRID HAVERSINE-GENETIC GEOGRAPHICAL MODELLING FOR AN OPTIMAL WEB SERVICE SELECTION/RECOMMENDATION.
- Authors
VAITHEKI, Kanagaraj; JOSEPH, Suresh
- Abstract
Web Service selection and recommendation becomes major aspects of service providing the community. Quality of Service prediction through Collaborative Filtering (CF) or Matrix Factorization (MF) is the key process in WS recommendation. Few or absence of QoS records (cold start), integration of QoS factors (response time and throughput), extensible models utilization and inaccurate similarity measures are the major problems in CF/MF approaches. This paper focuses on integration of similarity values with the geographical distance and QoS values (minimum response time and maximum throughput) to provide efficient mapping relationship (user-service and service-user). Initially, the geographical model is constructed by Haversine distance formulation based on latitude/longitude values. The Genetic Algorithm (GA) based recommendation list construction provides best WS as the responses to user query. The experimental validation of proposed work regarding response time, memory space, computation time, precision, coverage and number of patterns mined against existing methods assure effectiveness in QoS prediction.
- Publication
Economic Computation & Economic Cybernetics Studies & Research, 2019, Vol 53, Issue 4, p225
- ISSN
0424-267X
- Publication type
Academic Journal
- DOI
10.24818/18423264/53.4.19.14