Event extraction and prediction method based on time sequence event and semantic background

Abstract

The invention discloses an event extraction and prediction method based on a time sequence event and a semantic background, and the method comprises the following steps: (S1) carrying out the real-time collection of event data information through image collection, video collection and semantic background collection; (S2) the collected video stream is stored, and then video stream data information is obtained; meanwhile, a time sequence event stream is generated through the obtained video stream data information, and a timestamp is recorded; meanwhile, converting the acquired data information into a short text sample, and recording a short text sample label; (S3) extracting data features, performing data stream information feature extraction on video stream data information through an image recognition model, and performing data information analysis on a short text sample through a constructed classifier; (S4) constructing a prediction event prediction model and a semantic background model to realize data prediction; and (S5) outputting a prediction result. According to the method, extraction and prediction of time sequence events and semantic background events can be realized, and the event prediction capability is improved.

Description

technical field [0001] The invention relates to the field of time prediction and evaluation, and more specifically relates to an event extraction and prediction method based on time series events and semantic background. Background technique [0002] Event prediction has become a hot and difficult point of application in the field of computer vision in recent years. With the rapid development of computer technology, storage technology and network technology, as well as the continuous updating of various digital devices and mobile terminal devices, the data volume of various data information is exploding. The speed keeps growing. There are following technical defects in the prior art: [0003] (1) There is no standard for time prediction. When analyzing event information, the time flow is mixed and chaotic. It is impossible to obtain time series events from existing event data information, and the extraction accuracy is low, which leads to poor time evaluation and prediction...

AI Technical Summary

Problems solved by technology

[0003] (1) The prediction of time lacks standards. When analyzing event information, the time flow is mixed and chaotic. It is impossible to obtain time series events from existing event data information, and the extraction accuracy is low, which leads to poor time evaluation and prediction capabilities.

[0004] (2) Lack of rules, rule-based methods and statistical-based methods

In actual use, the existing dialogue recognition system is not very capable of recognizing user intentions, and there are often situations where the user cannot answer the user because the user's intention cannot be judged, or the answer is wrong or the answer is repeated, making the dialogue built by the dialogue system difficult. The content is too limited and the user experience is not high

Images