News text event and time extraction and standardization system for event tracking

Abstract

The invention aims to provide a news text event and time extraction and standardization system for event tracking. The news text event and time extraction and standardization system comprises a data acquisition and processing module (01), a news text preprocessing module (02), an event and time entity extraction module (03), a time standardization module (04) and a time axis establishment module (05); the invention comprises the following steps: firstly, carrying out clustering, part-of-speech tagging and other preprocessing on texts, then carrying out sub-event extraction and standardizationon the processed texts, carrying out similarity detection on standardized sub-events, carrying out time extraction and standardization on non-repetitive events, and finally, inserting new sub-events into an event time axis to which the new sub-events belong so as to complete continuous tracking of the events. When the time is normalized, the selection of the primitive time is not only limited to the current text, but is continuously associated with the preorder event of the event, so that the primitive time acquired by the method is more accurate.

Description

technical field [0001] The invention belongs to the technical field of event tracking, and in particular relates to a news text event, time extraction and normalization system for event tracking. Background technique [0002] With the rapid development of natural language processing, the recognition, extraction and reasoning of event information play an important role in text understanding. Especially in news event texts, the requirements for timeliness are high. From the beginning, duration, disposal to the end of an event, what happened at each point in time can establish an event tracking timeline for the event, which is very important for understanding the development of the situation, summarizing the event, and analyzing and summarizing the event afterwards. [0003] Defects and insufficiencies of the existing technology: At present, due to the complexity of Chinese expressions, the difficulty of semantic understanding, and the difficulty of text processing, most studi...

AI Technical Summary

Problems solved by technology

[0004] The problem to be solved by the present invention is that it is difficult to grasp the event primitive time with longer duration, and the problem of time extraction and standardization difficulty; the purpose of the invention is to provide a kind of news text event, time extraction and standardization system for event tracking. Carry out preprocessing such as clustering and part-of-speech tagging on the text, then perform sub-event extraction and normalization on the processed text, perform similarity detection on the normalized sub-events, perform time extraction and normalization on non-repeated events, and finally extract new sub-events Insert the timeline of the event to complete the continuous tracking of the event

Images