Entity matching method and device thereof

Abstract

The invention provides an entity matching method and a device thereof. The entity matching method comprises the steps as follows: obtaining training text information, performing word segmentation on the training text information to obtain an entity lexicon, constructing an entity word vector matrix according to frequency that every two entity words in the entity lexicon appear simultaneously in training text information, obtaining a target entity word mapped by a to-be-matched entity from the entity lexicon, obtaining a row vector corresponding to the target entity word from the word vector matrix, obtaining candidate column vectors corresponding to other word vectors except the column vector corresponding to the target entity word in the word vector matrix, calculating cosine similarity between a row vector corresponding to the target entity word and the candidate column vector, and determining an entity matched with the to-be-matched entity according to the calculated cosine similarity. Therefore, the entity matching efficiency can be improved.

Description

technical field [0001] The present invention relates to the technical field of data analysis, in particular to a method and device for entity matching. Background technique [0002] With the continuous development of social media, social media has gradually become the main way for people to obtain information. More and more people choose to refer to the information in social media platforms to formulate target plans. Based on the influence of various factors, the target plan is not Only, when the first target solution cannot be implemented, how to efficiently seek similar alternatives has become an urgent problem to be solved. For example, content related to celebrities and other entities has a high degree of discussion on social media platforms. When selecting actors for film and television, variety shows, or brand spokesperson promotion, while determining the number one candidate, it is also necessary to determine some Stars that are similar to the first-ranked candidate ...

AI Technical Summary

Problems solved by technology

In this method of entity matching, due to the different storage formats of user information in different social media platforms, it is difficult to analyze the data information obtained across platforms during the matching process. For example, information in different formats cannot be matched, resulting in a waste of matching resources. Or matching errors, making the matching results less reliable, resulting in lower efficiency of entity matching

Images