Mass disturbance warning method and system applied to short texts

Abstract

The invention discloses a mass disturbance warning method and system applied to short texts. The method comprises the following steps of automatically structuring a domain knowledge database applied to mass disturbances; combined with the domain knowledge database, performing local structured extraction and online classification on the short texts to screen out mass disturbance texts with potential risks from mass short texts; combined with the domain knowledge database, performing overall structured extraction and online clustering on identified short texts, and according to whether the number of short texts contained in every cluster exceeds a preset threshold value, determining whether to perform timely alarming. The mass disturbance warning method applied to the short texts has the advantages of fully integrating domain background knowledge, context, shallow semantic expression and deep semantic computation, achieving collaborative analysis and predication of mass disturbances and improving the timeliness, accuracy and recall rate of screening, tracking and alarming of the mass disturbances.

Description

technical field [0001] The present invention relates to the field of information security, and more specifically, relates to an early warning method and system for short-text group events. Background technique [0002] Our society is moving toward a stage where mass incidents frequently occur, such as the mass incidents that successively occurred in Lhasa, Tibet, Urumqi, Xinjiang, Bachu, Yunnan, and Kunming, Yunnan, etc., which not only caused tragic casualties, but also brought huge economic losses and severe disasters. social influence. In order to avoid or minimize the occurrence of mass incidents, the establishment of a reasonable and effective early warning and monitoring mechanism has begun to receive attention and attention. This requires the collection of relevant information and data reflecting public opinion trends, timely identification of potential mass conflicts and elimination of warning information, and access to information is the premise and source of crisi...

AI Technical Summary

Problems solved by technology

[0004] At present, some research works give early warning based on whether the data contains the sensitive words of interest. The preparation of the sensitive lexicon often requires manual participation and regular updates, and this method ignores the influence of the context. In fact, specific words are different in different contexts. For example, "walking" can be a sensitive word for calling collective events, but in many contexts, "walking" only represents a form of people's leisure or exercise

There are also research works that propose to automatically extract context features from data and train a "bag of words model" for description and prediction. This model assumes that the feature words are independent of each other and does not consider word order and syntactic features, which seriously affects the accuracy of early warning results. For example, "People's riots affect social stability", although the sentence contains the sensitive words "people" and "riots", it is not intentional to call an illegal assembly; Matching the category information obtained as a method of judging sensitive events or the same event may lead to information inconsistency, redundancy, etc.

It is also easy to see from the actual test results that there is still room for further improvement in the accuracy and comprehensiveness of information screening, tracking and early warning

Images