Product Category Recommendation System Using Markov Model

Abstract

Product recommendation is an intelligence system that is widely used in many industries. In order to recommend products to target customers efficiently, system designers must understand the characteristics of products and customers. In cosmetic industry, position and image of products influence to the customer’s decision. Some customers may stick with the same products, but some of them prefer like to keep trying new things. Thus, it is not easy to promote accurately cosmetics products to the users. Cosmetics advertisements usually come in a form of product lines or categories not just one item. If the marketers can narrow down the group of users who will buy in their preferred product lines, it will reduce cost and increases efficiency of that advertising especially in email marketing since it will not include non-target users. This paper presents a technique for category recommendation using clustering techniques and Markov model. With clustering, it leads the model to be more optimized and calibrated. Markov model can help to scope interested product category of each customer. This system also helped reduce advertising irrelevant products to the users.