AN EFFICIENT MARKETING POLICY RECOMMENDATION SYSTEM BASED ON USER CENTRIC SIMILARITY SEARCH USING REVERSE TOP-K QUERY

Sindhu B Jigali, Nirmala C R

Abstract: Information administration can be characterized as advancement and implementation of strategies, practices and systems. The advancement procedure comprises of recognizing likeness among information products as major task. Similitude measurements like Euclidian separation and cosine comparability are utilized to decide resemblance between information products and is figured in light of their properties. An effective advertisement marketing approach can be built up by not just thinking about the credits to process similitude among information products yet additionally client inclinations and conclusion on information about products. The set up arrangement focuses on the particular group of onlookers with bunch of best Data products. The closeness checking framework as a rule utilizes top-k question that profits the k Data products with the best rank for a specific client. The displayed inverted top-K inquiry output in the arrangement of powerful information products where the information products have a place with their top-k group. Jaccard constant is utilized to perform comparability calculations among the subsequent arrangements of the turnaround top-k questions, where as it additionally figures the min and max bound on the client driven closeness of information products. ?-similitude and m-closest neighbor inquiries productively register likeness among information about products of the invert top-k result.

Keywords: ?-similitude, m-closest nearest neighbors, top-k

DOI: https://doi.org/10.15623/ijret.2018.0707025