Event extraction method and system

Abstract

The invention provides an event extraction method. The method includes obtaining original corpora, labeling elements in sentences in the form of [label-element], dividing the marked corpus into a training set and a test set; mapping the training set into a vector by a pre-training language model; obtaining word embedding vectors, inputting the word embedding vector into a neural network model; outputting sequence label information by the neural network model; establishing a loss function based on the sequence label information, evaluating the event extraction model by using the test set to finally obtain an event extraction model, adjusting the training set and the test set structure for multiple times, selecting the event extraction model with the best evaluation result as an optimal model, and inputting a new original corpus into the optimal model to obtain an event extraction result; the event extraction method provided by the invention belongs to a supervised neural network extraction method, is combined with the fine-grained contextualized word vector in the field, and more conforms to the scene of event extraction in the professional field.

Description

technical field [0001] The invention belongs to the field of event extraction, in particular to an event extraction method and system. Background technique [0002] Event extraction is to extract the events that the user is interested in from the text describing the event information and present them in a structured form, so that events can be extracted from massive natural texts. At present, there are relatively few studies based on transformer fault event extraction. However, the effective use of a large number of transformer fault event records and experience is an important way to realize transformer fault intelligent diagnosis and reasoning. [0003] Existing event extraction methods are based on pattern matching and machine learning. Among them, pattern matching in the professional field needs to define a large number of templates for event recognition and extraction; methods based on traditional machine learning generally convert event extraction problems into classi...

AI Technical Summary

Problems solved by technology

[0004] The event extraction method based on pattern matching, based on a certain pattern (context environment), matches the sentence to be extracted with the existing template. The construction of the pattern requires expert knowledge in the field, and it is manually established, and the cost of manpower and time is high. , and the portability is poor, transplanting from one field to another is equivalent to re-establishing; the method of event extraction based on machine learning, one is to divide event extraction into two stages of entity extraction and event discrimination, the error of named entity recognition It will affect the discrimination of events and bring cumulative errors; second, for specific fields, it is often necessary to construct a large number of artificial features, and the process of feature selection is very costly. As the complexity of the model increases, the maintainability will change. getting worse

Images