Research On User Portrait Of Y Series Products Based On User Evaluation Dat

User profiles,as a collection of virtual user images that can describe user needs,interests,or personalized preferences,have become a hot research topic for enterprises and universities around the world,attracting widespread attention from the business and academic communities.This article analyzes the difficulties encountered in constructing user profiles for the YC industry and proposes corresponding solutions.Based on user evaluation data for Y series products in the YC industry,a user profile model(UPM)for Y series products is proposed.The main research content and innovative points of the paper are summarized as follows:1.A pre trained language model(PLM)was introduced to address the issue of difficulty in processing application comments or insufficient detail when constructing user profiles in evaluation data.A neural network model has been trained for mining user comment information for Y-series products,which can effectively extract user comment dimensions,viewpoint words,and emotional tendencies from comment text data,and construct user feedback dimensions and focus dimensions based on this.2.A UPM model for multi-dimensional user portraits of Y-series products is proposed.A more comprehensive user portrait is constructed from the four dimensions of user’s basic dimension,interaction dimension,feedback dimension and focus dimension.In the construction of interaction dimension and feedback dimension,the objective weighting method is introduced to quantify user attributes,making the portrait more detailed,and finally realizing visualization.3.Based on the UPM model,this article constructs a user profile from two specific application scenarios in the YC industry,applying different management perspectives to construct a user group profile,achieving visualization,and providing application ideas.The research ideas and methods of user portrait proposed in the paper,as well as the user personal portrait model constructed,are applicable to two different application scenarios in the YC industry,and can be expanded according to the actual application scenarios.At the same time,the above research results can also be extended to the research and application of user portraits with similar data structures or user information fields.