Method and device for constructing character relation graph and storage medium

Abstract

The invention discloses a method and device for constructing a figure relation graph and a storage medium, and the method comprises the steps: obtaining group chat data which comprises time information and text information; dividing the group chat data into one or more group chat segments according to the time information and the text information; extracting topic categories of the group chat segments, and classifying the group chat segments according to the topic categories to obtain one or more topic data, the topic data including one or more group chat segments; analyzing all group chat segments contained in the topic data to obtain a character relationship of the members, the character relationship including an active degree of the members, an interaction strength between the members and a social relationship between the members; and according to the character relationship of the members, constructing a character relationship graph of the members. According to the method and the device for constructing the character relation graph, the key chat content, the member information and the member relation can be further mined.

Description

technical field [0001] The present invention relates to the technical field of knowledge graphs, in particular to a method, device and storage medium for constructing a character relationship graph. Background technique [0002] In recent years, the mobile Internet has developed rapidly. As of December 2020, the number of Internet users in China has reached 989 million, of which mobile Internet users accounted for 99.7%, and instant messaging apps accounted for 99.2% of mobile Internet users. Group chat has the characteristics of low cost, group gathering and rapid dissemination, and has become a tool for criminals to publish false information and conduct online fraud. Therefore, group chat data has important analytical value. In the field of electronic data forensics analysis, quickly finding and locating key topics from massive mobile phone data, and at the same time effectively constructing the relationship between the people involved in the topic can often bring key prog...

AI Technical Summary

Problems solved by technology

At the same time, some methods (Li Gang, Wang Xiao, Guo Yang. Research on the internal relationship of WeChat groups based on the co-occurrence of members[J]. Data Analysis and Knowledge Discovery, 2018, 2(11): 54-63.) (Li Gang, Wang Xinping ,Ba Zhichao. Analysis of Conversational Network Structure and User Interaction Behavior in WeChat Groups[J].Intelligence Theory and Practice,2018,41(10):124-130). It is difficult to quickly and effectively collect evidence for key topics in relational positioning analysis

[0007] Secondly, although the word embedding method based on the pre-trained model can obtain a good key topic classification effect, the contribution of each word to the text classification is not equivalent, and there is a problem of redundant vocabulary features.

[0008] Finally, the traditional method still has the problem of only focusing on the topic and not being able to construct the relationship between the characters more accurately. It often lacks the construction of the relationship between the characters participating in the same sub-topic, which makes it difficult to find and locate key information in the case.

Images