User portrait extraction method based on text information

Abstract

The invention relates to the technical field of computer information processing, in particular to a text information-based user portrait extraction method, which mainly comprises the following steps of: 1, collecting text information; 2, judging whether the text information can be used for training and analysis of a computer for preliminary screening or not, and obtaining text sentences ; step 3,labeling the text sentences obtained in the step 2; 4, preprocessing the text sentences obtained in the step 3, extracting related data, and removing irrelevant words; and 5, constructing a text feature vector, and describing text information features through chi-square test and tfidf means. Compared with a traditional user portrait discovery system only based on rules, the human use efficiency can be effectively improved; the human cost of text information extraction is greatly reduced on the premise that the accuracy is guaranteed; and it can be guaranteed that the user portrait is obtainedefficiently on line. The purpose of efficiently and accurately extracting the user portrait in the text is achieved.

Description

technical field [0001] The invention relates to the technical field of computer information processing, in particular to a method for extracting user portraits based on text information. Background technique [0002] With the rapid development of information technology today, more and more fields are beginning to use machine learning technology to replace the complicated and repetitive work that consumes manpower and material resources in traditional industries. For example, in online shopping websites, voice assistants or conversational robots are used to solve common problems that customers may encounter, or the transportation department uses computer vision technology to recognize car license plates. The use of machine learning technology can not only effectively reduce production costs, but also ensure a high accuracy rate. [0003] In the actual industrial production process, extracting text information and marking the possible semantic information categories of a give...

AI Technical Summary

Problems solved by technology

However, it is often inefficient to filter texts by selecting keywords directly through labelers, and directly using the keyword list as a filtering condition may lead to problems such as high accuracy rate and low recall rate, or high recall rate and insufficient accuracy rate. Analyzing text using keywords alone is not enough

Images