Fashion recommendation has gained more interest leads to the developed of numerous E-commerce applications. Contemporary approaches generally suggest fashion products to the consumers based on the textual description provided. Comparing to the conventional product recommendations, fashion recommendations possess peculiar features such as time sensitive, present in different media, requires different kinds of data like size, body shape, and so on. Generally, the images of the products which cover a massive amount of information are usually not considered in the recommendation process. This paper presents an efficient Recommendation system (RS) for fashion products by employing the histogram based model. The presented histogram based RS (HRS) is particularly developed for image based queries to recommend identical products. By using the derived histograms, the images relevant to the given query are retrieved. On testing the presented HRS on the benchmark fashion image dataset, the HRS exhibited superior results over the compared methods in a significant way.
Volume 11 | 06-Special Issue
Pages: 249-259