Entity relationship joint extraction method based on transfer learning

Abstract

The invention discloses an entity relationship joint extraction method based on transfer learning. The method specifically comprises the following steps: taking a Chinese information extraction data set as a data source, preprocessing input sentences, using a Bert pre-training model, inputting a vector of an embedding layer into an encoder, acquiring a coding sequence, transmitting a word vector into a fully-connected Dense layer and a sigmoid activation function to obtain a coding vector of a main entity, transmitting the coding vector of the main entity to a fully-connected Dense network, predicting a guest entity and a relationship type, and combining with the main entity to finally obtain a triad. According to the method, transfer learning is applied to the entity-relationship joint extraction problem of a Chinese text, the triad can be directly modeled, the triad information is extracted from an unstructured text, and the relationship extraction efficiency and accuracy are remarkably improved.

Description

technical field [0001] The invention relates to the technical field of natural language processing, in particular, to a method for joint entity relation extraction based on migration learning. Background technique [0002] The extraction methods of entities and their relationships are mainly divided into serial extraction methods and joint extraction methods. Among them, the joint extraction method can integrate the information between entities and their relationships. With the success of deep learning in NLP tasks, neural networks are also widely used in entity and relational fact extraction. In 2016, "End-to-end relation extraction of sequences and tree structures based on LSTM" (Proceedings of the 54th Annual Meeting of the Association of Computational Linguistics) proposed an end-to-end model based on neural network to reduce the manual extraction of features. The implementation process separates the extraction of entities and their relationships resulting in informati...

AI Technical Summary

Problems solved by technology

In 2016, "LSTM-based Sequence and Tree Structure End-to-End Relational Extraction" (Proceedings of the 54th Annual Conference of the Association for Computational Linguistics) proposed an end-to-end model based on neural networks in order to reduce the work of manual feature extraction. The implementation process separates the extraction of entities and their relationships, resulting in information redundancy

For example, Bert proposed by Google first uses large-scale unsupervised data to pre-train the neural network model, and then uses the target data to fine-tune the model to adapt to the current task. In terms of Chinese data, research on entity-relationship joint extraction based on transfer learning is still in progress. very scarce

Images