Cigarette Brand Recommendation Based on Consumer Modeling
Dong Han, Peijiang Liu, Xiaotian Zhang, Keliang Jia · 2019
Based on the consumers' consumption records, the paper extracted the features which reflected consumers' preferences, calculated the weights of different eigenvalues of each feature based on the idea of TF-IDF algorithm, constructed a consumer's consumption preference model, and then extracted the same features and established the cigarette brand feature vector. A method for calculating consumer's preference similarity was defined. A cigarette brand recommendation algorithm based on consumer's preference was implemented by calculating the similarity between the cigarette brand feature vector and the consumer's preference feature. The experimental results showed that the algorithm achieved a high acceptance rate in the cigarette brand recommendation to the consumers.