Topic embedding based on network link and document content, document representation method

Abstract

The invention discloses a subject embedding and document representation method based on network link and document content. By introducing topological link information in document network, the problemof fuzzy semantic terms (such as synonyms) affecting topic in topic embedding model can be solved effectively. Using the link information, combining the document content and links in the data with probability graph model, optimizing the parameters of the model by variational inference method, the negative influence caused by fuzzy semantics is basically solved, and more accurate topic embedding and document representation are obtained. The method of the invention combines the link relation and the content information in the document network, The problem of inefficient topic embedding caused byfuzzy semantics (such as polysemy) in the existing topic embedding model is effectively improved. The probability graph model is established to make the method more interpretable, and the variationalexpectation maximization algorithm is used to make parameter updating efficient, convergence time short, and can be applied to large-scale networks.

Description

technical field [0001] The invention belongs to machine learning, complex network, and natural language processing, and proposes a new topic model to improve topic embedding (that is, to represent topics in low-dimensional space) to improve the low-dimensional representation of documents in document networks. Topic embedding of web links and document content, document representation methods. Background technique [0002] In the field of natural language processing, the topic model has many applications, such as: mapping documents to low-dimensional topic spaces, that is, using the topic distribution of documents to represent documents. The traditional topic model ignores the correlation between topics. In order to represent the correlation between topics, in recent years, in the topic model, the method of topic embedding has been proposed, which is mainly used to represent the relationship between topics. At present, most methods of obtaining document representation through...

AI Technical Summary

Problems solved by technology

However, the currently proposed method of obtaining document representation through topic embedding only utilizes the content information of the document (that is, the words in the document) to obtain the desired results, and there are a large number of fuzzy semantics (such as synonyms) in the content of the document, which is affected by fuzzy The impact of semantics can easily lead to poor topic embedding and document representation

Images