Fashion-based customer analysis for retail stores: A hybrid intelligent system proposal

Customer analysis in online and physical stores is a current significant approach used to decide customers' profiles and develop a selling strategy. With the help of these approaches, stores can decide on their target customers and products quite easily and plan and prepare optimal or suitable processes in online stores such as customer profile information, shopping history, following products, visiting frequency, etc. However, some difficulties are experienced to make such kinds of analyses at retail stores since they do not have such kind similar information. This thesis provides a detailed literature review covering analysis of stated topics in many terms; the problems and difficulties encountered in the field, current studies achieved, possible solutions, etc. Within the context of the thesis, a hybrid intelligent system based on artificial intelligence techniques was designed and developed to analyze customers respecting data privacy in terms of fashion in retail stores. Edge and cloud computing solutions were also used to support hybrid solution in the developed intelligent system having customer information as age, gender, ethnicity, emotion, clothes, styles, and colors and to detect in retail stores. Test results of the developed models have shown that the system works successfully to achieve the proposed tasks successfully and properly throughout the day, 80% and 66% success rates are achieved in offline and online tests. 93% and 100% accuracies were also obtained from the models of customer reidentification and face detection algorithms, respectively, on the online store tests respecting privacy. Clothes detection, clothes' color detection, and style detection were also successfully achieved with the scores of 75%, 65% and 57%, respectively. The proposed system and obtained results have shown that the system introduced in this study might help researchers not only in this specific fashion field but also other fields in stores to provide more facilities and opportunities.