Model training method and device, entity extraction method and device and storage medium

Abstract

The invention discloses a model training method, an entity extraction method, a model training device, an entity extraction device and a storage medium, which are used for solving the technical problem of low accuracy of entity extraction in public security texts in the prior art. The model training method comprises the following steps: performing word segmentation and part-of-speech tagging on each historical public security text in a historical public security text set; counting vocabularies in each word segmentation result in the tagged historical public security text set, the frequency of occurrence in each part-of-speech and the frequency of occurrence in adjacent positions of each part-of-speech to obtain corresponding word frequency probabilities and word use probabilities; generating a word vector of each vocabulary, and combining the word vector with the corresponding word frequency probability and word use probability to form a feature vector of the corresponding vocabulary; dividing all the feature vectors into a training set and a test set, wherein the training set is used for training the conditional random field model, and the test set is used for testing the trained conditional random field model; and stopping the training until the accuracy rate of a test result reaches a set threshold value to obtain the trained conditional random field model.

Description

technical field [0001] The invention relates to the field of public security information processing, in particular to a model training method, an entity extraction method, a device and a storage medium. Background technique [0002] In the field of public security, when handling various cases, investigators usually need to consult criminal investigation records and past case records related to the current case, and analyze these records to form a relevant knowledge map to analyze the current case. , Digging clues, stringing together cases, etc., in order to quickly solve the case. [0003] However, with the passage of time, more and more information about cases and persons involved in the public security system has been recorded. After finding the records related to the current case from the massive data, the investigators need to extract key elements from these case records ( Such as task, time, location, time, etc.), and establish the relationship between key elements, th...

AI Technical Summary

Problems solved by technology

[0007] The present invention provides a method for model training, a method for entity extraction, a device, and a storage medium to solve the technical problem that the accuracy of entity extraction in public security documents is not high in the prior art

Images