Open AccessOniomania Similarity Search with Recommended System
Author(s) -
K.Karpaga Priyaa*,
C. Saranya,
PA. Yogaalakshmi,
M. Haripriya
Publication year - 2019
Publication title -
international journal of recent technology and engineering
Language(s) - English
Resource type - Journals
ISSN - 2277-3878
DOI - 10.35940/ijrte.c6851.098319
Subject(s) - similarity (geometry) , ranking (information retrieval) , computer science , feature (linguistics) , product (mathematics) , information retrieval , rank (graph theory) , collaborative filtering , preference , recommender system , data mining , k nearest neighbors algorithm , nearest neighbor search , independence (probability theory) , artificial intelligence , mathematics , statistics , linguistics , philosophy , geometry , combinatorics , image (mathematics)
Users shopping online prefer to quickly access products that they are in need of among a list of similar products. The similarity metrics in use, only contemplates on the product attributes which doesn’t always meet the user expectations. The current system relies on solving this by performing Θ-similarity and r-nearest neighbor, where product similarity is considered only if it satisfies a user-preference list, by hitting the reverse top-k queries results. However, the products expressed here are generally based on the user’s preference, which can break the user independence of the product. To conquer these drawbacks, we use a more advanced system which represents feature based products such as Usercentric Model, where we have two phases. At first, we gather opinion based data about the feature, and subsequently these feature similarities are ranked. The ranking is predicted by collaborative filtering recommendation algorithm. Next, the rank generated from the above algorithm will be compared with existing data to get the top-k best product by performing No-Random Access algorithm (NRA) in second phase. Through this users are granted the top