CN113822069A - Emergency early warning method and device based on meta-knowledge and electronic device - Google Patents

Emergency early warning method and device based on meta-knowledge and electronic device Download PDF

Info

Publication number
CN113822069A
CN113822069A CN202111095062.1A CN202111095062A CN113822069A CN 113822069 A CN113822069 A CN 113822069A CN 202111095062 A CN202111095062 A CN 202111095062A CN 113822069 A CN113822069 A CN 113822069A
Authority
CN
China
Prior art keywords
event
target
time period
meta
knowledge
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202111095062.1A
Other languages
Chinese (zh)
Other versions
CN113822069B (en
Inventor
王丽宏
贺敏
李晨
郭舒
盛傢伟
钟盛海
黑一鸣
孙睿
范越
李倩
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
National Computer Network and Information Security Management Center
Original Assignee
National Computer Network and Information Security Management Center
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by National Computer Network and Information Security Management Center filed Critical National Computer Network and Information Security Management Center
Priority to CN202111095062.1A priority Critical patent/CN113822069B/en
Publication of CN113822069A publication Critical patent/CN113822069A/en
Application granted granted Critical
Publication of CN113822069B publication Critical patent/CN113822069B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/279—Recognition of textual entities
    • G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/30—Semantic analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/40—Processing or translation of natural language
    • G06F40/42—Data-driven translation
    • G06F40/49—Data-driven translation using very large corpora, e.g. the web
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Artificial Intelligence (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Evolutionary Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Biophysics (AREA)
  • Software Systems (AREA)
  • Mathematical Physics (AREA)
  • Computing Systems (AREA)
  • Molecular Biology (AREA)
  • Biomedical Technology (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The application relates to an emergency early warning method, an emergency early warning device and an electronic device based on meta-knowledge, wherein the method comprises the following steps: acquiring a target keyword in target text data and the time sequence heat of a first event in a first time period, wherein the target text data is text data describing the first event, and the target keyword is a keyword used for describing feature information of the first event in the target text data; determining meta-knowledge of the first event according to the target keywords, wherein the meta-knowledge is used for indicating target event characteristics of the first event; determining the target heat of the first event in a second time period according to the meta-knowledge and the time sequence heat in the first time period, wherein the ending time of the first time period is earlier than the starting time of the second time period; and determining the occurrence probability of the second event in the second time period according to the meta-knowledge and the target heat. The method and the device solve the technical problem that the efficiency of event early warning is low.

Description

Emergency early warning method and device based on meta-knowledge and electronic device
Technical Field
The present application relates to the field of computers, and in particular, to an emergency early warning method and apparatus based on meta-knowledge, and an electronic apparatus.
Background
With the rapid development of social networks, the traditional paper media is directly revolutionized, and the efficiency and the enthusiasm of people for obtaining, paying attention to and participating in social events are greatly improved. Therefore, more people manufacture and distribute massive information by means of new media, information diversification is promoted, and people can conveniently screen information needed by the people, the information can be rapidly spread in the diversified information mode, uncontrollable spreading of certain information is easy to occur, the uncontrollable events are defined as network emergencies, the network emergencies can be generated by chain reaction events, and timely early warning of the events is widely concerned as an increasingly important research target in order to avoid chain reaction events generated by the network emergencies.
The existing event early warning method still stays in the manners of dictionary matching, information cluster-based detection, keyword association analysis and the like of events in the event emergent judgment, and the manners either depend on the data scale of sample data and are difficult to label a large amount of data, or depend on manual experience and do not mine the essential reasons of the events, so that the event early warning effect is poor, and the accurate early warning can not be performed on the events.
In view of the above problems, no effective solution has been proposed.
Disclosure of Invention
The application provides an emergency early warning method and device based on meta-knowledge and an electronic device, and aims to at least solve the technical problem that the efficiency of event early warning in the related technology is low.
According to an aspect of an embodiment of the present application, there is provided an emergency early warning method based on meta-knowledge, including: acquiring a target keyword in target text data and the time sequence heat of a first event in a first time period, wherein the target text data is text data describing the first event, and the target keyword is a keyword used for describing feature information of the first event in the target text data; determining meta-knowledge of the first event according to the target keywords, wherein the meta-knowledge is used for indicating target event characteristics of the first event; determining the target heat of the first event in a second time period according to the meta-knowledge and the time sequence heat in the first time period, wherein the ending time of the first time period is earlier than the starting time of the second time period; and determining the occurrence probability of a second event in a second time period according to the meta-knowledge and the target heat, wherein the second event is an event to be warned caused by the first event.
Optionally, determining meta-knowledge of the first event according to the target keyword comprises: generating a word vector of the target keyword by using a word vector generation model; generating a word vector set according to the word vectors of the target keywords, wherein the word vectors in the word vector set are stored according to a target sequence, and the target sequence is determined according to the position information of the target keywords in the text data; generating a first event vector of a first event by using a pre-trained event feature generation model and a word vector set, wherein the event feature generation model is obtained by training an initial event generation model by using a training sample set, the training sample set comprises a positive sample and a negative sample, and the negative sample is obtained by modifying the positive sample; and determining meta-knowledge of the first event according to the relationship between the first event vector of the first event and a plurality of second event vectors of a plurality of preset events stored in the preset event set.
Optionally, determining meta-knowledge of the first event according to a relationship between the first event vector of the first event and a plurality of second event vectors of a plurality of preset events stored in the preset event set comprises: determining the spatial relationship between the first event and a plurality of preset events according to the first event vector and the second event vector; performing event clustering on the first event and the plurality of preset events based on the spatial relationship to obtain a target event cluster containing the first event, wherein the target event cluster stores the first event and the preset events in a preset neighborhood range of the first event; and determining the average value of the event vectors in the target event class cluster as the meta-knowledge of the first event.
Optionally, the determining the target heat of the first event in the second time period according to the meta-knowledge and the time sequence heat in the first time period comprises: the method comprises the steps of obtaining heat values of a first event at a plurality of preset moments in a first time period, and a third event vector of the first event at each preset moment, wherein the first time period comprises the plurality of preset moments; weighting and calculating a third event vector of the first event at each preset moment by using first weight information and second weight information to obtain a fourth event vector of the first event in a first time period, wherein the first weight information is generated according to historical weight information by using a weight information generation model, and the second weight information is determined according to the similarity of the third event vector and the meta knowledge of the first event; and generating a target heat degree of the second time period according to the fourth event vector and the time sequence heat degree in the first time period by using a preset vector generation model.
Optionally, determining the probability of occurrence of the second event at the second time period according to the meta-knowledge and the target heat comprises: predicting an initial probability value of the event type of the first event in the first time period as a target type according to the meta-knowledge by using an event type prediction model; and calculating the initial probability value and the target heat degree by using a first formula to obtain the occurrence probability of the second event in the second time period.
Optionally, after determining the probability of occurrence of the second event for the second time period according to the meta knowledge and the target heat, the method further comprises: detecting whether the occurrence probability value of the second event meets a preset condition or not; generating a target notification message under the condition that the occurrence probability value of the second event meets a preset condition; and sending the target notification message to preset terminal equipment.
According to another aspect of the embodiments of the present application, there is also provided an emergency early warning apparatus based on meta-knowledge, including: the acquisition module is used for acquiring a target keyword in the target text data and the time sequence heat of the first event in a first time period, wherein the target text data is text data describing the first event, and the target keyword is a keyword which is used for describing feature information of the first event in the target text data; the first determining module is used for determining meta-knowledge of the first event according to the target key words, wherein the meta-knowledge is used for indicating target event characteristics of the first event; the second determining module is used for determining the target heat of the first event in the second time period according to the meta-knowledge and the time sequence heat in the first time period, wherein the ending time of the first time period is earlier than the starting time of the second time period; and the third determining module is used for determining the occurrence probability of a second event in a second time period according to the meta-knowledge and the target heat, wherein the second event is an event to be early-warned caused by the first event.
Optionally, the first determining module includes: a first generation unit configured to generate a word vector of the target keyword using a word vector generation model; a second generating unit configured to generate a word vector set based on word vectors of the target keywords, wherein the word vectors in the word vector set are stored in a target order, and the target order is determined based on position information of the target keywords in the text data; a third generating unit, configured to generate a first event vector of the first event using the pre-trained event feature generation model and the word vector set; the determining unit is used for determining the meta-knowledge of the first event according to the relation between the first event vector of the first event and a plurality of second event vectors of a plurality of preset events stored in the preset event set.
According to another aspect of the embodiments of the present application, there is also provided a storage medium including a stored program which, when executed, performs the above-described method.
According to another aspect of the embodiments of the present application, there is also provided an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the above method through the computer program.
In the embodiment of the application, target keywords in target text data and time sequence heat of a first event in a first time period are obtained, wherein the target text data is text data describing the first event, and the target keywords are keywords used for describing feature information of the first event in the target text data; determining meta-knowledge of the first event according to the target keywords, wherein the meta-knowledge is used for indicating target event characteristics of the first event; determining the target heat of the first event in a second time period according to the meta-knowledge and the time sequence heat in the first time period, wherein the ending time of the first time period is earlier than the starting time of the second time period; determining the occurrence probability of a second event in a second time period according to the meta-knowledge and the target heat, wherein the second event is a to-be-early-warning event caused by a first event, by acquiring a target keyword in target text data for describing the first event, the acquired target keyword is a keyword for describing feature information of the first event, determining meta-information for indicating the event feature of the first event according to the target keyword, thereby obtaining the essential feature of the first event, predicting the heat information of the event in the second time period after the first time period according to the meta-information of the first event and the time sequence heat of the first event in the first time period, and further determining the probability of the first event causing the early-warning event in the second time period according to the heat information of the event in the second time period and the meta-information of the event, the method achieves the purpose of describing keywords in text data describing the first event and the time sequence heat of the first event in the first time period and accurately predicting the probability of the early warning event caused by the first event in the second time period, thereby achieving the technical effect of improving the efficiency of event early warning and further solving the technical problem of low efficiency of event early warning.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without inventive exercise.
FIG. 1 is a schematic diagram of a hardware environment for a method of event probability determination according to an embodiment of the present application;
FIG. 2 is a flow chart of an alternative meta-knowledge based emergency alert method according to an embodiment of the present application;
FIG. 3 is a schematic diagram of an alternative heat prediction according to an embodiment of the present application;
FIG. 4 is a flow chart of an alternative emergency alert process according to an embodiment of the present application;
FIG. 5 is a schematic diagram of an alternative meta-knowledge based emergency alert device according to an embodiment of the present application; and
fig. 6 is a block diagram of a terminal according to an embodiment of the present application.
Detailed Description
In order to make the technical solutions better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only partial embodiments of the present application, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It should be noted that the terms "first," "second," and the like in the description and claims of this application and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the application described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
According to an aspect of the embodiments of the present application, an embodiment of a method for early warning of an emergency based on meta-knowledge is provided.
Alternatively, in this embodiment, the emergency early warning method based on meta-knowledge may be applied to a hardware environment formed by the terminal 101 and the server 103 as shown in fig. 1. As shown in fig. 1, the server 103 is connected to the terminal 101 through a network, which may be used to provide services (such as data query services, data calculation services, etc.) for the terminal or a client installed on the terminal, and a database may be provided on the server or separately from the server, and is used to provide data storage services for the server 103, and the network includes but is not limited to: the terminal 101 is not limited to a PC, a mobile phone, a tablet computer, and the like. The emergency early warning method based on meta-knowledge in the embodiment of the application may be executed by the server 103, or may be executed by the terminal 101, or may be executed by both the server 103 and the terminal 101. The terminal 101 executing the emergency warning device method based on meta-knowledge according to the embodiment of the present application may also be executed by a client installed thereon.
Fig. 2 is a flowchart of an alternative emergency early warning method based on meta-knowledge according to an embodiment of the present application, and as shown in fig. 2, the method may include the following steps:
step S202, acquiring a target keyword in target text data and the time sequence heat of a first event in a first time period, wherein the target text data is text data describing the first event, and the target keyword is a keyword used for describing feature information of the first event in the target text data;
step S204, determining meta-knowledge of the first event according to the target keywords, wherein the meta-knowledge is used for indicating the target event characteristics of the first event;
step S206, determining the target heat of the first event in a second time period according to the meta-knowledge and the time sequence heat in the first time period, wherein the ending time of the first time period is earlier than the starting time of the second time period;
and step S208, determining the occurrence probability of a second event in a second time period according to the meta-knowledge and the target heat, wherein the second event is an event to be warned caused by the first event.
Through the steps S202 to S208, by obtaining the target keyword in the target text data describing the first event, where the obtained target keyword is a keyword for describing feature information of the first event, and determining meta-information indicating an event feature of the first event according to the target keyword, so as to obtain an essential feature of the first event, and by predicting heat information of the first event in a second time period after the first time period according to the meta-information of the first event and a time sequence heat of the first event in the first time period, and further determining a probability that the first event causes an early warning event in the second time period according to the heat information of the event in the second time period and the meta-information of the event, the purpose of accurately predicting a probability that the first event causes an early warning event in the second time period according to the keyword in the text data describing the first event and the time sequence heat of the first event in the first time period is achieved, therefore, the technical effect of improving the efficiency of event early warning is achieved, and the technical problem that the efficiency of event early warning is low is solved.
In the technical solution provided in step S202, the target text data may be input by the target account, or may be generated in real time according to monitoring of a certain event.
Optionally, in this embodiment, the target text data may be a text that changes with the development of the target event in the first time period, such as: target text data in the development status of a certain company, a company is created by Zhang III, the company is named as a first company, the target text data is that the first company is created by Zhang III, the Zhang III is purchased by Liquan after a period of time and is renamed to be a second company, and the updated text data is that the first company is purchased by Liquan and is renamed to be a second company.
Optionally, in this embodiment, the first time period may include, but is not limited to, a plurality of preset sub-time periods or a plurality of time points, and the time-series heat is a heat of each sub-time period or each event point.
Optionally, in this embodiment, the time-series popularity may be, but is not limited to, obtained by monitoring related content in real time by using a detection tool, for example, to detect a browsing amount, a click rate, a forwarding amount, a topic occurrence rate of a detection target event on social software, and the like of the related content of a certain website, which is not limited in this embodiment.
Optionally, in this embodiment, the length of the first time period may be flexibly set according to requirements, for example, the length of the first time period may be set to 1 minute, 5 minutes, 1 day, 1 week, and so on.
Optionally, in this embodiment, the number of the target keywords may be one or more, and the target keywords may include, but are not limited to, keywords related to a time trigger word, keywords of an event type, keywords of an event participant, keywords of a role of a participant, and the like, which is not limited in this embodiment.
In the technical solution provided in step S204, the meta knowledge may be obtained by performing correlation calculation on the target keyword, or may be obtained by comparing the target keyword with an existing event, and the meta knowledge is used to describe an essential reason for the occurrence of the target event.
In the technical solution provided in step S206, the length of the second time period may be set according to requirements, for example, the length of the second time period may be set to, but is not limited to, 1 second, 2 seconds, 1 minute, 5 minutes, and the like, and this solution is not limited thereto.
Optionally, in this embodiment, the method for determining the target heat degree in the second time period may include, but is not limited to, generating the target heat degree in the first time period according to the meta knowledge and the time series heat degree in the first time period by using a pre-trained heat degree generation model, and may also be obtained by calculating the meta knowledge and the time series heat degree by using a formula.
In the technical solution provided in step S208, the manner of determining the second event occurrence probability in the second time period may include, but is not limited to, generating the second event occurrence probability according to the meta knowledge and the target heat using a trained model, and may also be obtained by calculating the meta knowledge and the target heat using a preset formula, which is not limited in this embodiment.
Optionally, in this embodiment, the second event is an early warning time at a second time period caused by the first event, such as: when the first event is that Zhang III donates to the public welfare tissue for a plurality of times in 6 and 7 months of the year and the amount is huge, the second event pays attention to the fact that Zhang III donates and the amount is huge for most people in a certain period of time in the future.
As an alternative embodiment, determining meta-knowledge of the first event based on the target keyword comprises:
s11, generating a word vector of the target keyword by using the word vector generation model;
s12, generating a word vector set according to the word vectors of the target keywords, wherein the word vectors in the word vector set are stored according to a target sequence, and the target sequence is determined according to the position information of the target keywords in the text data;
s13, generating a first event vector of a first event by using a pre-trained event feature generation model and a word vector set, wherein the event feature generation model is obtained by training an initial event generation model by using a training sample set, the training sample set comprises a positive sample and a negative sample, and the negative sample is obtained by modifying the positive sample;
s14, determining meta-knowledge of the first event according to a relationship between the first event vector of the first event and a plurality of second event vectors of a plurality of preset events stored in the preset event set.
Optionally, in this embodiment, the word vector generation model is obtained by performing word vector generation training using news corpora in daily news.
Optionally, in this embodiment, the roles that words play in different positions in the target text information are different (for example, "zhang san catches lie four," zhang san "is a subject and is an object for initiating an action," lie four "is an object and is an object for accepting an action," catch "is an action triggered by zhang san), and the storage order of the target keywords is determined according to the positions of the target keywords in the target text data, so as to obtain the word vector set.
Optionally, in this embodiment, the event feature generation model is obtained by training the initial event feature generation model using a training sample set, where the training sample set includes a positive sample and a negative sample, the positive sample may be an event whose preset event vector and word vector of the corresponding keyword are known, and the negative sample may be obtained by randomly adjusting and replacing any keyword of the positive sample.
Through the steps, the text data is used for describing the first event, word vectors of the target keywords in the text data are obtained, the word vectors are sorted and stored according to the positions of the corresponding target keywords in the text to obtain the word vector set, so that the first event vector generated by the event vector generation model according to the word vector set is more accurate, the meta-knowledge of the first event is determined through the relation between the first event and the event vectors of the stored preset events in the preset event set, and the determined meta-information of the first event is more accurate and reliable.
As an alternative embodiment, determining meta-knowledge of the first event according to a relationship between the first event vector of the first event and a plurality of second event vectors of a plurality of preset events stored in the preset event set comprises:
s21, determining the spatial relationship between the first event and a plurality of preset events according to the first event vector and the second event vector;
s22, performing event clustering on the first event and the preset events based on the spatial relationship to obtain a target event cluster containing the first event, wherein the target event cluster stores the first event and the preset events in a preset neighborhood range of the first event;
and S23, determining the average value of the event vectors in the target event class cluster as the meta-knowledge of the first event.
Optionally, in this embodiment, the event vector is multidimensional, and the vectors with different dimensions further determine the position of the event in the space, so as to determine the position relationship between the first event and the preset event in the space.
Optionally, in this embodiment, the neighborhood may be randomly set according to the use requirement.
Optionally, in this embodiment, clustering may begin from any data point that has never been visited, the neighborhood of change points is divided by distance, if there are a sufficient number of points in the domain, the clustering process begins and the store becomes the first point in the new cluster, otherwise the point will be marked as noise; for the first point in the new cluster, the points in its neighborhood will become part of the same cluster, and the process will have all points in the domain belonging to the same cluster, and repeat the above operations for all new points that have just been added to the cluster group, so that all points in space have cluster groups.
Through the steps, the target event class cluster of the first event is determined according to the position relation of the first event and the event vector of the preset event in the space, the preset event in the neighborhood range of the first event is stored in the target event class cluster, and then the points in the target event class cluster can be considered as the events most similar to the intrinsic characteristics of the first event, so that the meta-knowledge of the first event is determined, and the obtained meta-knowledge is closer to the intrinsic characteristics of the first event.
As an alternative embodiment, the determining the target heat of the first event in the second time period according to the meta-knowledge and the timing heat in the first time period comprises:
s31, obtaining heat values of a plurality of preset moments of a first event in a first time period and a third event vector of the first event at each preset moment, wherein the first time period comprises the plurality of preset moments;
s32, weighting and calculating a third event vector of the first event at each preset time by using first weight information and second weight information to obtain a fourth event vector of the first event in a first time period, wherein the first weight information is generated according to historical weight information by using a weight information generation model, and the second weight information is determined according to the similarity of the third event vector and the meta knowledge of the first event;
and S33, generating a target heat degree of the second time period according to the fourth event vector and the time sequence heat degree in the first time period by using a preset vector generation model.
Optionally, in this embodiment, the preset time is preset according to a requirement, for example, 5 seconds, 10 seconds, 20 seconds, etc. may be set as one preset time.
Optionally, in this embodiment, the first weight information may be generated according to a conventional attention mechanism, and the first weight information may also be generated by calculating historical weight information.
Optionally, in this embodiment, before performing weighting and calculation, normalization processing may be performed on the first weight information and the second weight information, so that the first weight information and the second weight information are in the same dimension, and the reliability of the calculation result is improved.
FIG. 3 is a schematic diagram of an alternative heat prediction according to an embodiment of the present application, as shown in FIG. 3: x0 and x1 … … xi are respectively the heat value of each time interval in the first time period, h0 and h1 … … hi are respectively the event characteristics of the first event at each moment in the first time period, and the event is continuously changed along with the change of time, so the vector value of the event is continuously changed, and alpha isi_sIs a first weight value, alpha, determined under the conventional attention mechanismi_mBased on similarity of event characteristics at each time and meta-knowledge of the first eventAnd Hi is a characteristic value weighted value and a heat weighted value which are obtained by respectively weighting and summing the event characteristics and the heat values at all the moments in the time period from 0 to i according to the first weight information and the second weight information, and then a preset vector generation model is used for generating the target heat of the event at the moment i + 1.
As an alternative embodiment, the determining the probability of the occurrence of the second event in the second time period according to the meta-knowledge and the target heat degree comprises:
s41, predicting an initial probability value of the event type of the first event in the first time period as the target type according to the meta-knowledge by using the event type prediction model;
and S42, calculating the initial probability value and the target heat degree by using the first formula to obtain the occurrence probability of the second event in the second time period.
Optionally, in this embodiment, the event type prediction model may be a binary classification model of a multi-layer neural network, and the event prediction model may be obtained by training an initial event type prediction model using a sample set, where the sample set includes positive samples and negative samples, the positive samples and the negative samples are both labeled samples, the positive samples may be emergencies whose meta-knowledge and event type are known, and the negative samples are meta-knowledge of non-emergencies in daily news and corresponding event types.
Optionally, in this embodiment, the first formula may be, but is not limited to being
Figure BDA0003268914910000131
Figure BDA0003268914910000132
Is the probability of occurrence of a second event, P, for a second period of timeemergencyIs an initial probability value, P, that the event type of the first event is a target type during a first time periodhotFor the target heat, α and β are preset weight values.
As an alternative embodiment, after determining the probability of occurrence of the second event for the second time period according to the meta-knowledge and the target heat, the method further comprises:
s51, detecting whether the target probability value meets a preset condition;
s52, generating a target notification message under the condition that the target probability value meets the preset condition;
and S53, sending the target notification message to the preset terminal equipment.
Optionally, in this embodiment, the preset condition may be that the target probability value is within the target interval range, or that the target probability value satisfies a certain set threshold, which is not limited by the present solution.
Fig. 4 is an alternative flow chart of emergency warning according to an embodiment of the present application, as shown in fig. 4:
s401, as semi-structured information (text data describing an event) combined by literal symbols, the event needs to be represented as a computable numerical value. By carrying out word vector training in a large amount of news corpora, the scheme can convert event information such as event types and event trigger words into word vectors for word representation. Then, an event eiThe keywords corresponding to the event elements can be represented by splicing, and the word vectors of the keywords corresponding to the event elements are arranged and stored according to a predetermined sequence, wherein the predetermined sequence is determined according to the positions of the event elements in the text data, namely [ wtype:wtrigger:wsubject:wobject]Wherein w istypeWord vectors for keywords corresponding to event types, wtriggerWord vectors for event-triggered words, wsubjectIs a keyword feature value corresponding to the subject information, wobjectThe method is characterized in that the characteristic value of a keyword corresponding to object information, the event type and the trigger word are only one, and if more than one information such as an object, a subject and the like exists, the mean value of all object and subject information representations is taken.
S402, because the number of emergencies is small and the labeling is difficult, less data can be used for training and analyzing. Therefore, the scheme designs a set of self-supervision tasks to pre-train the event representation so as to improve the quality of the input of the model (namely the event representation). The event element classification refers to classifying different event elements according to four categories (event type, event trigger word, event argument and argument type); the event authenticity detection means that a false event is generated as a negative example by randomly replacing information such as event trigger words, objects, event categories and the like, and the false event and a real emergency form a sample set to perform classification judgment; the event heat degree prediction refers to the detection of the matching degree based on the event elements and the heat degree; the event element detection refers to extraction training based on an original event text and a common sequence labeling model.
S403, in order to extract the event burst meta-knowledge, it is necessary to find the intrinsic cause of the event in the labeled emergency. According to the invention, the semantic distance of the event representation is defined as the similarity of the burst reasons, so that the essential reasons of the event burst are converted into the discovery of the event cluster. In the present invention, clustering may begin with any data point that has not been visited, with the neighborhood of the point divided by a distance ε. If there are a sufficient number of points in the neighborhood, the clustering process begins and the point becomes the first point in the new cluster. Otherwise, the point will be marked as noise. In both cases, the point is marked as "access". For the first point in the new cluster, the points within distance ε will become part of the same cluster. The process causes all points in the neighborhood of ε to belong to the same cluster and repeats at all new points just added to the cluster group. The process of the first two steps will repeat until all points in the cluster are determined, and all points in the vicinity of the cluster are visited and labeled. When all point clustering is completed, a new unvisited point is retrieved and processed, resulting in new cluster or noise discovery. By repeating this process, all points will be marked as visited, and clustering is complete.
S404, obtaining n class clusters of the event representations in step S403, and obtaining meta-event representations (i.e. the essential cause and the meta-knowledge of the event burst) of each class cluster by taking the average value of all event representations in the class cluster.
S405, after the meta-knowledge of the emergency is extracted, the emergency detection can be performed. The invention is a semi-structure for subsequent inputThe same initialization and pre-training is performed for the chemometric events as for the event samples in the training set. And then, introducing a multilayer neural network as a two-classification model, wherein positive examples of sources are marked emergencies and the extracted meta-knowledge in the text, and negative examples of sources are non-emergencies in massive news texts. Because of the dichotomy, the final result to be predicted by the model is only two cases, and the probability obtained by prediction of the invention for each category is pemergencyAnd p1-emergency。
S406, removing the sudden detection of the event, wherein the future heat degree of the event is also an important judgment index for early warning or not. Wherein the heat of an event refers to the share of the event element in all events over different time periods. By forming the heat of the events in different time periods into a time sequence, the prediction of the heat of the event at the future moment is converted into a classic time sequence prediction problem. Because the development of event heat is linked with event elements and mutual relations, the invention designs the realization of taking the meta-knowledge of the incident burst as an attention model, and the invention divides the time period according to the heat to obtain the elements of the incident in the known time period and generates corresponding incident representation xiAnd encoding the time sequence of 0-i time period to generate Hi(characteristic values of events over time period 0 to i) and hoti(Heat of event in time period 0 to i) by adding HiAnd hotiAs the input of the time period, the invention designs a corresponding time sequence decoder to realize the heat prediction of the i +1 time period.
S407, after the sudden detection and the future time prediction of the event are finished, the early warning of the sudden event can be directly carried out, the probability that the event is the sudden event and the heat degree prediction value of the i +1 time period are calculated by using a first formula to obtain the probability that whether the event triggers the early warning time in the i +1 time period or not, wherein the first formula is
Figure BDA0003268914910000161
pemergencyIs that the event is a burstProbability value of occurrence, PhotAnd alpha and beta are preset weight values of the sudden event detection and the heat degree predicted value.
It should be noted that, for simplicity of description, the above-mentioned method embodiments are described as a series of acts or combination of acts, but those skilled in the art will recognize that the present application is not limited by the order of acts described, as some steps may occur in other orders or concurrently depending on the application. Further, those skilled in the art should also appreciate that the embodiments described in the specification are preferred embodiments and that the acts and modules referred to are not necessarily required in this application.
Through the above description of the embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but the former is a better implementation mode in many cases. Based on such understanding, the technical solutions of the present application may be embodied in the form of a software product, which is stored in a storage medium (e.g., ROM/RAM, magnetic disk, optical disk) and includes instructions for enabling a terminal device (e.g., a mobile phone, a computer, a server, or a network device) to execute the method of the embodiments of the present application.
According to another aspect of the embodiment of the present application, there is also provided an emergency warning device based on meta-knowledge, which is used for implementing the emergency warning method based on meta-knowledge. Fig. 5 is a schematic diagram of an alternative meta-knowledge based emergency alert device according to an embodiment of the present application, and as shown in fig. 5, the device may include:
the obtaining module 52 is configured to obtain a target keyword in the target text data and a time sequence heat of the first event in a first time period, where the target text data is text data describing the first event, and the target keyword is a keyword in the target text data, where the keyword is used to describe feature information of the first event;
a first determining module 54, configured to determine meta-knowledge of the first event according to the target keyword, where the meta-knowledge is used to indicate a target event characteristic of the first event;
a second determining module 56, configured to determine a target heat of the first event in a second time period according to the meta-knowledge and a timing heat in the first time period, where an end time of the first time period is earlier than a start time of the second time period;
and a third determining module 58, configured to determine, according to the meta-knowledge and the target heat, an occurrence probability of a second event in a second time period, where the second event is an event to be warned caused by the first event.
It should be noted that the obtaining module 52 in this embodiment may be configured to execute the step S202 in this embodiment, the first determining module 54 in this embodiment may be configured to execute the step S204 in this embodiment, the second determining module 56 in this embodiment may be configured to execute the step S206 in this embodiment, and the third determining module 58 in this embodiment may be configured to execute the step S208 in this embodiment.
It should be noted here that the modules described above are the same as the examples and application scenarios implemented by the corresponding steps, but are not limited to the disclosure of the above embodiments. It should be noted that the modules described above as a part of the apparatus may operate in a hardware environment as shown in fig. 1, and may be implemented by software or hardware.
Through the module, the technical problem of low efficiency of event early warning can be solved, and the technical effect of improving the efficiency of event early warning is achieved.
Optionally, the first determining module includes: a first generation unit configured to generate a word vector of the target keyword using a word vector generation model; a second generating unit configured to generate a word vector set based on word vectors of the target keywords, wherein the word vectors in the word vector set are stored in a target order, and the target order is determined based on position information of the target keywords in the text data; a third generating unit, configured to generate a first event vector of a first event using a pre-trained event feature generation model and a word vector set, where the event feature generation model is obtained by training an initial event generation model using a training sample set, the training sample set includes a positive sample and a negative sample, and the negative sample is obtained by modifying the positive sample; the determining unit is used for determining the meta-knowledge of the first event according to the relation between the first event vector of the first event and a plurality of second event vectors of a plurality of preset events stored in the preset event set.
Optionally, the determining unit is configured to: determining the spatial relationship between the first event and a plurality of preset events according to the first event vector and the second event vector; performing event clustering on the first event and the plurality of preset events based on the spatial relationship to obtain a target event cluster containing the first event, wherein the target event cluster stores the first event and the preset events in a preset neighborhood range of the first event; and determining the average value of the event vectors in the target event class cluster as the meta-knowledge of the first event.
Optionally, the second determining module includes: the device comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring heat values of a first event at a plurality of preset moments in a first time period and a third event vector of the first event at each preset moment, and the first time period comprises the plurality of preset moments; the first calculation unit is used for weighting and calculating a third event vector of the first event at each preset moment by using first weight information and second weight information to obtain a fourth event vector of the first event in a first time period, wherein the first weight information is generated according to historical weight information by using a weight information generation model, and the second weight information is determined according to the similarity of the third event vector and the meta knowledge of the first event; and the fourth generation unit is used for generating the target heat of the second time period according to the fourth event vector and the time sequence heat in the first time period by using a preset vector generation model.
Optionally, the third determining module includes: the prediction unit is used for predicting an initial probability value of the event type of the first event in the first time period as a target type according to the meta-knowledge by using an event type prediction model; and the second calculating unit is used for calculating the initial probability value and the target heat degree by using the first formula to obtain the occurrence probability of the second event in the second time period.
Optionally, the apparatus further comprises: the detection module is used for determining the occurrence probability of a second event in a second time period according to the meta knowledge and the target heat degree and then detecting whether the occurrence probability value of the second event meets a preset condition; the generating module is used for generating a target notification message under the condition that the occurrence probability value of the second event meets a preset condition; and the sending module is used for sending the target notification message to the preset terminal equipment.
It should be noted here that the modules described above are the same as the examples and application scenarios implemented by the corresponding steps, but are not limited to the disclosure of the above embodiments. It should be noted that the modules described above as a part of the apparatus may be operated in a hardware environment as shown in fig. 1, and may be implemented by software, or may be implemented by hardware, where the hardware environment includes a network environment.
According to another aspect of the embodiment of the present application, there is also provided a server or a terminal for implementing the above emergency early warning method based on meta-knowledge.
Fig. 6 is a block diagram of a terminal according to an embodiment of the present application, and as shown in fig. 6, the terminal may include: one or more processors 601 (only one of which is shown), a memory 603, and a transmission device 605. as shown in fig. 6, the terminal may further include an input-output device 607.
The memory 603 may be configured to store software programs and modules, such as program instructions/modules corresponding to the method and apparatus for emergency warning based on meta-knowledge in the embodiment of the present application, and the processor 601 executes various functional applications and data processing by running the software programs and modules stored in the memory 603, that is, implements the method for emergency warning based on meta-knowledge. The memory 603 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 603 may further include memory located remotely from the processor 601, which may be connected to the terminal through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The above-mentioned transmission device 605 is used for receiving or sending data via a network, and may also be used for data transmission between a processor and a memory. Examples of the network may include a wired network and a wireless network. In one example, the transmission device 605 includes a Network adapter (NIC) that can be connected to a router via a Network cable and other Network devices to communicate with the internet or a local area Network. In one example, the transmission device 605 is a Radio Frequency (RF) module, which is used for communicating with the internet in a wireless manner.
Among them, the memory 603 is used to store an application program, in particular.
The processor 601 may call the application stored in the memory 603 through the transmission device 605 to perform the following steps: acquiring a target keyword in target text data and the time sequence heat of a first event in a first time period, wherein the target text data is text data describing the first event, and the target keyword is a keyword used for describing feature information of the first event in the target text data; determining meta-knowledge of the first event according to the target keywords, wherein the meta-knowledge is used for indicating target event characteristics of the first event; determining the target heat of the first event in a second time period according to the meta-knowledge and the time sequence heat in the first time period, wherein the ending time of the first time period is earlier than the starting time of the second time period; the method and the device for early warning the emergency based on the meta knowledge and the scheme of the electronic device are provided. The method comprises the steps of obtaining a target keyword in target text data for describing a first event, wherein the obtained target keyword is a keyword for describing feature information of the first event, determining meta-information for indicating event features of the first event according to the target keyword so as to obtain essential features of the first event, predicting heat information of the event in a second time period after the first time period according to the meta-information of the first event and the time sequence heat of the first event in the first time period, determining the probability of the first event triggering an early warning event in the second time period according to the heat information of the event in the second time period and the meta-information of the event, and achieving the purpose of accurately predicting the probability of the first event triggering the early warning event in the second time period according to the keyword in the text data for describing the first event and the time sequence heat of the first event in the first time period, therefore, the technical effect of improving the efficiency of event early warning is achieved, and the technical problem that the efficiency of event early warning is low is solved.
Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments, and this embodiment is not described herein again.
It can be understood by those skilled in the art that the structure shown in fig. 6 is only an illustration, and the terminal may be a terminal device such as a smart phone (e.g., an Android phone, an iOS phone, etc.), a tablet computer, a palm computer, and a Mobile Internet Device (MID), a PAD, etc. Fig. 6 is a diagram illustrating a structure of the electronic device. For example, the terminal may also include more or fewer components (e.g., network interfaces, display devices, etc.) than shown in FIG. 6, or have a different configuration than shown in FIG. 6.
Those skilled in the art will appreciate that all or part of the steps in the methods of the above embodiments may be implemented by a program instructing hardware associated with the terminal device, where the program may be stored in a computer-readable storage medium, and the storage medium may include: flash disks, Read-Only memories (ROMs), Random Access Memories (RAMs), magnetic or optical disks, and the like.
Embodiments of the present application also provide a storage medium. Alternatively, in this embodiment, the storage medium may be a program code for executing an emergency alert method based on meta knowledge.
Optionally, in this embodiment, the storage medium may be located on at least one of a plurality of network devices in a network shown in the above embodiment.
Optionally, in this embodiment, the storage medium is configured to store program code for performing the following steps: acquiring a target keyword in target text data and the time sequence heat of a first event in a first time period, wherein the target text data is text data describing the first event, and the target keyword is a keyword used for describing feature information of the first event in the target text data; determining meta-knowledge of the first event according to the target keywords, wherein the meta-knowledge is used for indicating target event characteristics of the first event; determining the target heat of the first event in a second time period according to the meta-knowledge and the time sequence heat in the first time period, wherein the ending time of the first time period is earlier than the starting time of the second time period; and determining the occurrence probability of a second event in a second time period according to the meta-knowledge and the target heat, wherein the second event is an event to be warned caused by the first event.
Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments, and this embodiment is not described herein again.
Optionally, in this embodiment, the storage medium may include, but is not limited to: a U-disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a removable hard disk, a magnetic or optical disk, and other various media capable of storing program codes.
The above-mentioned serial numbers of the embodiments of the present application are merely for description and do not represent the merits of the embodiments.
The integrated unit in the above embodiments, if implemented in the form of a software functional unit and sold or used as a separate product, may be stored in the above computer-readable storage medium. Based on such understanding, the technical solution of the present application may be substantially implemented or a part of or all or part of the technical solution contributing to the prior art may be embodied in the form of a software product stored in a storage medium, and including instructions for causing one or more computer devices (which may be personal computers, servers, network devices, or the like) to execute all or part of the steps of the method of the embodiments of the present application.
In the above embodiments of the present application, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
In the several embodiments provided in the present application, it should be understood that the disclosed client may be implemented in other manners. The above-described embodiments of the apparatus are merely illustrative, and for example, a division of a unit is merely a division of a logic function, and an actual implementation may have another division, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, units or modules, and may be in an electrical or other form.
Units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The foregoing is only a preferred embodiment of the present application and it should be noted that those skilled in the art can make several improvements and modifications without departing from the principle of the present application, and these improvements and modifications should also be considered as the protection scope of the present application.

Claims (10)

1. An emergency early warning method based on meta-knowledge is characterized by comprising the following steps:
acquiring a target keyword in target text data and the time sequence heat of a first event in a first time period, wherein the target text data is text data describing the first event, and the target keyword is a keyword which is used for describing feature information of the first event in the target text data;
determining meta-knowledge of the first event according to the target keyword, wherein the meta-knowledge is used for indicating a target event characteristic of the first event;
determining a target heat degree of the first event in a second time period according to the meta-knowledge and the time sequence heat degree in the first time period, wherein the ending time of the first time period is earlier than the starting time of the second time period;
and determining the occurrence probability of a second event in the second time period according to the meta-knowledge and the target heat, wherein the second event is an event to be warned caused by the first event.
2. The method of claim 1, wherein determining the meta-knowledge of the first event based on the target keyword comprises:
generating a word vector of the target keyword by using a word vector generation model;
generating a word vector set according to the word vectors of the target keywords, wherein the word vectors in the word vector set are stored according to a target sequence, and the target sequence is determined according to position information of the target keywords in the text data;
generating a first event vector of the first event by using a pre-trained event feature generation model and the word vector set, wherein the event feature generation model is obtained by training an initial event generation model by using a training sample set, the training sample set comprises positive samples and negative samples, and the negative samples are obtained by modifying the positive samples;
determining the meta-knowledge of the first event according to a relationship between the first event vector of the first event and a plurality of second event vectors of a plurality of preset events stored in a preset event set.
3. The method of claim 2, wherein determining the meta-knowledge of the first event according to the relationship between the first event vector of the first event and a plurality of second event vectors of a plurality of the preset events stored in the preset event set comprises:
determining a spatial relationship between the first event and a plurality of preset events according to the first event vector and the second event vector;
performing event clustering on the first event and the preset events based on the spatial relationship to obtain a target event cluster containing the first event, wherein the first event and the preset events in a preset neighborhood range of the first event are stored in the target event cluster;
determining an average of event vectors in the target event class cluster as the meta-knowledge of the first event.
4. The method of claim 1, wherein determining the target heat for the first event over the second time period based on the meta-knowledge and the temporal heat for the first time period comprises:
acquiring heat values of a first event at a plurality of preset moments in a first time period and a third event vector of the first event at each preset moment, wherein the first time period comprises the plurality of preset moments;
weighting and calculating the third event vector of the first event at each preset moment by using first weight information and second weight information to obtain a fourth event vector of the first event in the first time period, wherein the first weight information is generated according to historical weight information by using a weight information generation model, and the second weight information is determined according to the similarity of the third event vector and the meta knowledge of the first event;
generating the target heat of the second time period according to the fourth event vector and the time sequence heat in the first time period by using a preset vector generation model.
5. The method of claim 1, wherein determining the probability of the second event occurring during the second time period based on the meta-knowledge and the target heat comprises:
predicting an initial probability value of the event type of the first event as a target type in the first time period according to the meta-knowledge by using an event type prediction model;
and calculating the initial probability value and the target heat degree by using a first formula to obtain the occurrence probability of the second event in the second time period.
6. The method of claim 1, wherein after determining the probability of the second event occurring during the second time period based on the meta-knowledge and the target heat, the method further comprises:
detecting whether the occurrence probability value of the second event meets a preset condition or not;
generating a target notification message under the condition that the occurrence probability value of the second event meets a preset condition;
and sending the target notification message to preset terminal equipment.
7. An emergency early warning device based on meta-knowledge, comprising:
the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring a target keyword in target text data and the time sequence heat of a first event in a first time period, the target text data is text data describing the first event, and the target keyword is a keyword in the target text data for describing feature information of the first event;
a first determining module, configured to determine meta-knowledge of the first event according to the target keyword, where the meta-knowledge is used to indicate a target event feature of the first event;
a second determining module, configured to determine a target heat of the first event in a second time period according to the meta-knowledge and the time-sequence heat in the first time period, where an end time of the first time period is earlier than a start time of the second time period;
and a third determining module, configured to determine, according to the meta-knowledge and the target heat, an occurrence probability of a second event in the second time period, where the second event is an event to be warned caused by the first event.
8. The apparatus of claim 7, wherein the first determining module comprises:
a first generating unit configured to generate a word vector of the target keyword using a word vector generation model;
a second generating unit, configured to generate a word vector set according to the word vectors of the target keywords, where the word vectors in the word vector set are stored according to a target order, and the target order is determined according to position information of the target keywords in the text data;
a third generating unit, configured to generate a first event vector of the first event using a pre-trained event feature generation model and the word vector set;
a determining unit, configured to determine the meta-knowledge of the first event according to a relationship between the first event vector of the first event and a plurality of second event vectors of a plurality of preset events stored in a preset event set.
9. A storage medium, characterized in that the storage medium comprises a stored program, wherein the program when executed performs the method of any of the preceding claims 1 to 6.
10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor executes the method of any of the preceding claims 1 to 6 by means of the computer program.
CN202111095062.1A 2021-09-17 2021-09-17 Sudden event early warning method and device based on meta-knowledge and electronic device Active CN113822069B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111095062.1A CN113822069B (en) 2021-09-17 2021-09-17 Sudden event early warning method and device based on meta-knowledge and electronic device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111095062.1A CN113822069B (en) 2021-09-17 2021-09-17 Sudden event early warning method and device based on meta-knowledge and electronic device

Publications (2)

Publication Number Publication Date
CN113822069A true CN113822069A (en) 2021-12-21
CN113822069B CN113822069B (en) 2024-03-12

Family

ID=78914796

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111095062.1A Active CN113822069B (en) 2021-09-17 2021-09-17 Sudden event early warning method and device based on meta-knowledge and electronic device

Country Status (1)

Country Link
CN (1) CN113822069B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116542238A (en) * 2023-07-07 2023-08-04 和元达信息科技有限公司 Event heat trend determining method and system based on small program

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103455639A (en) * 2013-09-27 2013-12-18 清华大学 Method and device for recognizing microblog burst hotspot events
CN108228808A (en) * 2017-12-29 2018-06-29 东软集团股份有限公司 Determine the method, apparatus of focus incident and storage medium and electronic equipment
CN109299258A (en) * 2018-09-18 2019-02-01 平安科技（深圳）有限公司 A kind of public sentiment event detecting method, device and equipment
CN109582785A (en) * 2018-10-31 2019-04-05 天津大学 Emergency event public sentiment evolution analysis method based on text vector and machine learning
WO2019184217A1 (en) * 2018-03-26 2019-10-03 平安科技（深圳）有限公司 Hotspot event classification method and apparatus, and storage medium
CN113343118A (en) * 2021-04-23 2021-09-03 东南大学 Hot event discovery method under mixed new media
CN113378565A (en) * 2021-05-18 2021-09-10 北京邮电大学 Event analysis method, device and equipment for multi-source data fusion and storage medium

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103455639A (en) * 2013-09-27 2013-12-18 清华大学 Method and device for recognizing microblog burst hotspot events
CN108228808A (en) * 2017-12-29 2018-06-29 东软集团股份有限公司 Determine the method, apparatus of focus incident and storage medium and electronic equipment
WO2019184217A1 (en) * 2018-03-26 2019-10-03 平安科技（深圳）有限公司 Hotspot event classification method and apparatus, and storage medium
CN109299258A (en) * 2018-09-18 2019-02-01 平安科技（深圳）有限公司 A kind of public sentiment event detecting method, device and equipment
CN109582785A (en) * 2018-10-31 2019-04-05 天津大学 Emergency event public sentiment evolution analysis method based on text vector and machine learning
CN113343118A (en) * 2021-04-23 2021-09-03 东南大学 Hot event discovery method under mixed new media
CN113378565A (en) * 2021-05-18 2021-09-10 北京邮电大学 Event analysis method, device and equipment for multi-source data fusion and storage medium

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
刘巧玲;李劲;肖人彬;: "基于参数反演的网络舆情传播趋势预测――以新浪微博为例", 计算机应用, no. 5, pages 1419 - 1423 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116542238A (en) * 2023-07-07 2023-08-04 和元达信息科技有限公司 Event heat trend determining method and system based on small program
CN116542238B (en) * 2023-07-07 2024-03-15 和元达信息科技有限公司 Event heat trend determining method and system based on small program

Also Published As

Publication number Publication date
CN113822069B (en) 2024-03-12

Similar Documents

Publication Publication Date Title
US11487941B2 (en) Techniques for determining categorized text
CN109525595B (en) Black product account identification method and equipment based on time flow characteristics
Alfarrarjeh et al. Geo-spatial multimedia sentiment analysis in disasters
CN110163647B (en) Data processing method and device
Hunt et al. Monitoring misinformation on Twitter during crisis events: a machine learning approach
US20190392258A1 (en) Method and apparatus for generating information
WO2023124204A1 (en) Anti-fraud risk assessment method and apparatus, training method and apparatus, and readable storage medium
KR102399787B1 (en) Recognition of behavioural changes of online services
CN108885623A (en) The lexical analysis system and method for knowledge based map
Chen et al. Bert-log: Anomaly detection for system logs based on pre-trained language model
Kim et al. Anomaly pattern detection for streaming data
Win et al. Target oriented tweets monitoring system during natural disasters
CN112365171A (en) Risk prediction method, device and equipment based on knowledge graph and storage medium
Dou et al. Pc 2 a: predicting collective contextual anomalies via lstm with deep generative model
CN111177367A (en) Case classification method, classification model training method and related products
CN110209921B (en) Method and device for pushing media resource, storage medium and electronic device
CN116865994A (en) Network data security prediction method based on big data
CN116996325A (en) Network security detection method and system based on cloud computing
Zhu et al. CCBLA: a lightweight phishing detection model based on CNN, BiLSTM, and attention mechanism
Ma et al. Social media event prediction using DNN with feedback mechanism
CN110674300A (en) Method and apparatus for generating information
CN113822069A (en) Emergency early warning method and device based on meta-knowledge and electronic device
Ravishankar et al. Empirical Assessment and Detection of Suicide Related Posts in Twitter using Artificial Intelligence enabled Classification Logic
CN117729003A (en) Threat information credibility analysis system and method based on machine learning
CN105302844B (en) Internet surveillance method, apparatus and system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant