Social network comment generation method based on LSTM

Abstract

The invention discloses a social network comment generation method based on LSTM, and belongs to the technical field of social network comment generation. The social network comment generation methodand device solve the problem that a material library cannot be provided for public opinions due to the fact that an applied scene of an existing social network comment generation technology is too narrow and single. According to the method, an NLG technology based on LSTM learning is used, and the visual semantics of a sentence structure, the types of characters and each character are coded through the probability relation between the characters obtained through learning. Semantic and syntactic fusion is carried out on comment information to be expressed, and a vivid, smooth and variable high-quality comment text which is almost consistent with a social network is generated through methods such as specific word replacement in the later period. A favorable material corpus is provided for public opinion guidance, and a positive energy network environment is restored by spreading more real and trustworthy languages. The method can serve as a material corpus to be input into an existing public opinion guiding system, and is used for generating comments in the specific field of the social network.

Description

technical field [0001] The invention relates to an LSTM-based social network comment generation method, belonging to the technical field of social network comment generation. Background technique [0002] Nowadays, online social networking platforms have greatly promoted the life and communication of netizens. People and things around the world are closely connected because of the Internet, and people's participation in network events is getting higher and higher, resulting in numerous social network comments. Comments represent a discourse, a voice, and a reflection of ideology, with concise text sentences, clear intentions, and diverse structures, making them an ideal place to test automatic text generation techniques. Some users express their opinions and positions by posting posts on a certain hot event, or agree, or neutral, or deny, or it is an Internet rumor driven by a kind of interest. This patent locks the social network on the Twitter platform, collects user-gene...

AI Technical Summary

Problems solved by technology

[0008] In order to understand the problem that the existing social network comment generation technology is too narrow and single to provide a material library for public opinion, the present invention further proposes an LSTM-based social network comment generation method

Images