Designing an e-commerce recommender system based on collaborative filtering using a data mining approach

An effective recommender system based on personality traits, demographics and behavior of customers in time context

“…In general, the main contributions of this research are: 1) proposing an accurate RS using personality traits, demographic characteristics and time-based purchasing behavior; 2) proposing an effective and comprehensive RS for a "clothing" online store; 3) improving the RS performance by solving the cold start issue using personality traits and demographic characteristics; 4) improving the scalability issue in RS through efficient k-nearest neighbors; 5) Mitigating the sparsity issue by using personality traits and demographic characteristics and also by densifying the user-item matrix and 6) improving the RS accuracy by solving the 1. Introduction A strong and accurate recommender system (RS) is one of the most important tools for the success and profitability of e-businesses (Khodabandehlou, 2019;Koohi and Kiani, 2017). One of the problems faced by the visitors and customers of commercial websites, especially clothing-related websites, is their exposure to a large number of similar and diverse products, which makes it difficult for them to choose one and also prolongs the buying process and sometimes confuses the customer and may finally cause them to leave the online stores without making a purchase.…”

“…CF assumes that customers who are similar to each other (i.e. they buy the same or similar products) also have similar interests and are like-minded (Das et al, 2014;Khodabandehlou, 2019). As a result, such systems attempt to predict interests and preferences of customers and, based on the similarity between them, recommend products of their interest to the related customers (Chen and He, 2009;Son, 2014); in order to do so, CF selects the subset of the target customer's nearest neighbors based on the similarity between them and recommends a list of items to the target customer according to the neighbors' purchases and also by considering the similarity rate between the purchases (Borr as et al, 2014;Nilashi et al, 2018).…”

“…CF, despite its success and widespread use, faces several issues, including cold start (Safoury and Salah, 2013;Wu and Chen, 2015), scalability (Inan et al, 2018;Sahu and Dwivedi, 2019), sparsity (Sahu and Dwivedi, 2019;Zhang et al, 2014) and instability of customer interest and preferences over time (Ding et al, 2006;Khodabandehlou, 2019). The cold start issue happens when the system still does not have sufficient information to make recommendations, i.e.…”

“…(Inan et al, 2018;Zhang et al, 2014). In addition to the above issues, there is a more serious issue faced by CF that is called "interest drift", which means that interests and needs of customers may change over time, due to the change of social, cultural and economical life conditions (Khodabandehlou, 2019;Mal et al, 2007). As CF tries to recommend products that are interesting for and relevant to the customer's needs based on their previous purchases, interest drift can be a serious challenge for it.…”

“…For example, a mother will no longer tend to buy baby clothes after her baby is grown up. In other words, the customer's purchase of goods at one time reflects their recognition and interest in the goods, which may fade or change over time (Khodabandehlou, 2019;Lee et al, 2008). Thus, it can be concluded that customers' interests and preferences are better reflected by their recent purchases than their older ones (Ding et al, 2006;Rafeh and Bahrehmand, 2012;Sen et al, 2015).…”