【Purpose/significance】The purpose of online shopping user interest portrait based on review information is to find the user

interest through user review information, so as to provide users with personalized product recommendation，etc..【

Method/process】This paper takes the evaluation of cosmetics on Taobao as the research object. Build a dictionary of user interests dictionary. Determine the user portrait dimension by the value attribute of the commodity, and do user interest profile analysis. The process of establishing a dic⁃tionary of user interests: First of all, the ROST CM 6 software is used to process the comment information to find word segmentation and delete the stop word. Then the word frequency analysis method is used to mine the vocabulary of interest. Finally, a dictionary was established with the "Vocabulary for Emotion Analysis (beta)" published by CNKI.【Result/conclusion】User portrait analysis is based on the user interest dictionary, and uses many tools and methods, such as emotion analysis technology, TAG Cloud and semantic net⁃work analysis technology, etc..【Innovation/limitation】 User Interest Dictionary is the innovation of this paper. The selection of soft⁃ware tools fully considers the advantages and limitations of different software tools and dictionaries. It is an exploration how to use data analysis tools. The algorithm based the user's interest profile is not done that push the goods to the individual users. It is a deficiency of this article.