基于文本挖掘增强的画像系统设计与实现

Design and Implementation of Portrait System Based on Text Mining Enhancement

基于用户的网络行为数据进行画像分析,有益于解决因海量信息导致的难以精准了解用户喜好及需求的问题。以用户在涉及眼科医疗领域的互联网浏览行为为分析对象,对用户阅读内容进行文本挖掘,通过词向量模型构建、特征选择、主题分类等环节,完成对阅读内容的画像构建。对用户不同网络行为进行挖掘分析,同时考虑用户兴趣的实时变化,将用户不同的网络行为映射为相应的权重,并通过衰减系数进行计算,完成对用户画像的动态刻画,从而最终得到完整准确的基于文本挖掘增强的画像系统。

Profile analysis based on users’ network behavior data is beneficial to solve the problem of difficult to accurately understand user preferences and needs caused by massive information. In this paper, the user’s Internet browsing behavior in the field of ophthalmology is used as the analysis object to conduct text mining on the user’s reading content.Through the construction of word vector model, feature selection, topic classification and other links, the profile construction of the reading content is completed. Mining and analysis of different network behaviors of users, taking into account the real-time changes of user interests, mapping different network behaviors of users into corresponding weights and calculating through attenuation coefficients to complete the characterization of user profile, so as to finally obtain a complete and accurate text-based mining into the enhanced profile system.