CN111143566A - Method for predicting hot event outbreak aiming at twitter - Google Patents
Method for predicting hot event outbreak aiming at twitter Download PDFInfo
- Publication number
- CN111143566A CN111143566A CN201911390362.5A CN201911390362A CN111143566A CN 111143566 A CN111143566 A CN 111143566A CN 201911390362 A CN201911390362 A CN 201911390362A CN 111143566 A CN111143566 A CN 111143566A
- Authority
- CN
- China
- Prior art keywords
- user
- tweet
- author
- relationship
- forwarded
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000000354 decomposition reaction Methods 0.000 claims description 7
- 238000004458 analytical method Methods 0.000 abstract description 8
- 230000008451 emotion Effects 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 11
- 230000003993 interaction Effects 0.000 description 5
- 230000002457 bidirectional effect Effects 0.000 description 2
- 241000984082 Amoreuxia Species 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000003012 network analysis Methods 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Primary Health Care (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to a method for predicting hot event outbreak aiming at a twitter, which is used for predicting whether a certain hot event in the twitter can cause large-scale propagation. The tweet comprises one or more topics, the tweet is classified according to the topics to form a data set, and the relationship between the topics and the tweet is a many-to-many relationship. And then combing the forwarding relation and the appointed forwarding relation in the tweet and forming a data set. Integrating the required tweets according to a certain topic needing to be predicted, building a user relation network, and decomposing and expressing the network. Finally, the decomposed network prediction is used. Compared with the traditional emotion analysis and user analysis methods, the method can comprehensively know the network state of the event and the range of the users from the perspective of the user relationship network composition of the whole network, can timely and quickly know the development trend of the current event, and can give a prediction result.
Description
Technical Field
The invention relates to a hot event outbreak prediction method based on an event propagation network, and belongs to the technical field of Internet public opinion analysis.
Background
With the rapid development and popularization of the internet and social media, twitter and microblog become important platforms for people to obtain important media and social public opinions of news current affairs, interpersonal interaction, self expression, social sharing and social participation by virtue of the characteristics of openness, terminal expansibility, content compactness, low threshold and the like. For some social phenomena or social events, people prefer to express their own opinions through social media such as twitter or microblog and participate in the discussion of the events.
The more people focus on an event, the more people who trigger the discussion, and the number of associated users increases, which eventually leads to the outbreak of the event as more and more users join the discussion of the event.
Disclosure of Invention
The method mainly aims at social media platforms such as twitter and microblog, realizes the prediction of each stage of an event through the research of a relation network formed by forwarding and @ between certain event tweet users, can give effective early warning information, and provides proper reference for public sentiment control.
The traditional public opinion analysis idea is mainly to analyze public opinions by methods such as emotion analysis and the like from the perspective of texts; or look at a large V of the origination history to predict the forwarding amount of new tweets, etc. The analysis method is relatively limited, cannot integrally and macroscopically reflect the overall situation of the network, and can analyze the change of the network from the angle of the change of the whole network by practical propagation network analysis so as to obtain whether a certain event becomes a hot event or has a tendency of outbreak.
In order to achieve the above object, the present invention adopts a technical solution as a method for predicting hot spot event outbreaks based on an event propagation network, as shown in fig. 1, the method is implemented as follows:
step 1: obtaining data in twitter real-time data stream
Step 2: classification of tweets
Obtaining the tweet from the step 1, classifying the tweet by using the words in the text, and establishing a tweet classification data set;
and step 3: pushing context relationship combing
There are four cases for a tweet:
in the first case, the text is originated from the original, and there is no forwarding or forwarding to the specified object;
in the second case, the text is originated from the original, and there is a case of forwarding to a specific object;
in the third situation, the text is transferred from other authors, and the text is not transferred to the specified object;
the fourth case, text going from other authors, exists for forwarding to a particular object.
In the above four cases, the first case has no association relationship between users, and therefore, does not process it; in the second case, a piece of information is stored in the database, namely, the specified relationship existing between the author and the forwarded object is recorded; in the third case, recording the forwarding relation between the author and the forwarded tweet author; in the fourth case, two records are created, the forwarding relation between the author of the text and the author to be forwarded and the designated relation between the author of the text and the specific object to be forwarded are recorded respectively, and finally, a user relation data set of the tweet is formed;
and 4, step 4: prediction
The method makes a prediction for whether a topic can cause large-scale propagation, and the prediction step is divided into three steps:
the first step is as follows: organizing a user relationship network for a topic
The text pushing topic data set is obtained in the step 2, and for a certain topic, text pushing data of all the topics can be searched in the data set according to the ID of the topic; according to the ID of the tweet, obtaining the user relationship tables of all the tweets at present from the user relationship data set obtained in the step 3;
the second step is that: expressing a topic network
The first step is to obtain a user relationship table, form a user relationship network according to the user relationship of the table, and decompose and express the user relationship network according to 13 seed graph models, namely [ l1, l2, …, l13], wherein li represents the number of the i-th seed graph model decomposed by the user relationship network;
step three: prediction
The change of a certain event in a certain period of the user relationship network is reflected on the change of the subgraphs, so that the period change of the number of the subgraphs is analyzed to predict whether the event has an outbreak trend in a period in the future.
In step 1, the data comprises ID (tweet ID), user _ ID (tweet author ID), user _ location (tweet author country), profile (author head address), screen _ name (author nickname), content (content of the tweet), and created _ at (tweet creation time);
if the piece of tweet is forwarded from other authors, the following fields are also included in the data: retweeted (forwarded tweed content), retweeted _ ID (ID of forwarded tweed), retweeted _ screen _ name (nickname of forwarded tweed author), retweeted _ user _ ID (user ID of forwarded tweed), retweeted _ user _ location (country of forwarded tweed author), retweeted _ created _ at (creation time of forwarded tweed); each tweet contains a plurality of topics, and the relationship between the topics and the tweets is a many-to-many relationship. The decomposition method of the user relationship network in the step 4 is a decomposition method mentioned by Batagelj V, Mrvar A in the paper A Subqualative triple center Algorithm for Large Sparse Networks with Small Maximum Degreee, and the method relates to 13 seed graph models for decomposing the network.
Advantageous effects
Compared with the traditional emotion analysis and user analysis methods, the method can comprehensively know the network state of the event and the range of the users from the perspective of the user relationship network composition of the whole network, can timely and quickly know the development trend of the current event, and can give a prediction result.
Drawings
FIG. 1a is a user relationship diagram;
FIG. 1b, 1c, 1d user relationship subgraphs;
FIGS. 2 and 13 are schematic diagrams of seed map models;
FIG. 3, method flow diagram.
Detailed Description
The invention adopts the technical scheme that the method for predicting the hot event outbreak based on the event propagation network comprises the following implementation steps:
step 1: obtaining data in twitter real-time data stream
Desensitized real-time tweet data may be received using an official API (application programming interface, hereinafter collectively referred to as API) through twitter's official website application developer account. The data contains ID (tweed), user _ ID (tweet author ID), user _ location (tweet author country), profile (author avatar address), screen _ name (author nickname), content (content of the piece of tweet), and created _ at (tweet creation time), and if the piece of tweet is forwarded from other authors, the data also contains the following fields: retweeted (forwarded tweed content), retweeted _ ID (ID of forwarded tweed), retweeted _ screen _ name (nickname of forwarded tweed author), retweeted _ user _ ID (user ID of forwarded tweed), retweeted _ user _ location (country of forwarded tweed author), retweeted _ created _ at (creation time of forwarded tweed). The following table is the messages received from the API of twitter:
from the above example, we can see that each tweet contains many topics, for example, the tweet contains two topics of # gilett janes and # yellowtests. Therefore, the topic is also a hidden attribute, a piece of tweet can contain a plurality of topics, and the relationship between the topic and the tweet is a many-to-many relationship. RT @ someone in this document refers to a tweet forwarded from a user nickname @ someone, indicating that the tweet is not originally authored by the author, the content before RT is what the author wants to say, and the content after RT @ someone is what the author forwards publishes. For purposes of this document, the author has only published "@ charles 134" below, with other content published by the forwarding author.
Step 2: classification of tweets
From step 1, we get the tweets, and we classify the tweets with the words in the text. And establishing a tweet classification data set.
If the contents of the tweet in step 1 are "@ charles134 RT @ ohboywyhashot: 13th week: # GiletsJanes # yellowshows project: # Paris-Onggoing projects in all cities in France-Mainstream media: …", the tweet contains two topics of # GiletsJanes and # Yellows, so we can build two classification records for the tweet. The procedure for establishing the record is as follows:
first, check whether there is ID of the topic in the database, if not, create the topic, and the topic ID is randomly generated.
| Topic name | Topic ID |
| GiletsJaunes | 6a7wdr |
| Yellowvests | 5a7gdw |
Secondly, two topic records of the tweet are created, and the record contents are as follows:
| tweet ID | Topic ID |
| 1095022875428425730 | 6a7wdr |
| 1095022875428425730 | 5a7gdw |
And step 3: pushing context relationship combing
There are four cases for a tweet:
in the first case, the text is originated from the original, and there is no forwarding or forwarding to the specified object;
in the second case, the text is originated from the original, and there is a case of forwarding to a specific object;
in the third situation, the text is transferred from other authors, and the text is not transferred to the specified object;
the fourth case, text going from other authors, exists for forwarding to a particular object.
In the first of the above four cases, there is no association between users, and therefore no processing is performed. In the second case, we will store a piece of information in the database, i.e. record the specified relationship between the author and the forwarded object. In the third case, we record the forwarding relationship between the author of this document and the author of the forwarded tweet. In the fourth case, we create two records, which record the forwarding relationship between the author and the author to be forwarded and the specified relationship between the author and the specific object to be forwarded. And finally forming a user relationship data set of the tweet.
For example, the content of the tweet in step 1 is: "@ charles134 RT @ ohboywhastoot: 13th week: # GiletsJaunes # Yellowtestets: # Paris-ingprotests in all cities in France-Mainstream media: …", the nickname of the author herein is Pammybarrett, since the API will not contain the user ID of the specified object, but twitter specifies that the whole net of the nickname is unique, so that it is feasible to establish the relationship between users by using the nickname of the user, and the steps are as follows:
first, we build a forwarding relationship between Pammybarrett and ohboywatasoot in the data, since the tweet is forwarded from an author called ohboywatasoot on a nickname.
We see in this context that the author forwarded to a user who is nickname charles134, so we establish a specified relationship in the database between pammmybarrett and charles 134.
We use the number 1 to represent the forwarding relationship and the number 2 to represent the designated relationship, then the information in the database is:
| tweet ID | Authors refer to | Relationship author | Relationships between | |
| 1095022875428425730 | Pammybarrett | ohboywhatashot | 1 | |
| 1095022875428425730 | | charles134 | 2 |
And 4, step 4: prediction
The amount of topics in twitter is very large, and the use scene is that whether a certain topic can cause large-scale propagation is predicted. The prediction step is divided into three steps:
the first step is as follows: organizing a user relationship network for a topic
In step 2, we obtain a tweet topic data set, and for a topic, we can retrieve tweet data of all the topics in the data set according to the ID of the topic. According to the ID of the tweet, the user relationship tables of all the tweets at present can be obtained from the user relationship data set obtained in the step 3, and the results of the relationship tables are as follows:
| authors refer to | Relationship author | Relationships between | |
| Pammybarrett | ohboywhatashot | 1 | |
| | charles134 | 2 |
The second step is that: expressing a topic network
In the first step, a user relation table is obtained, and a user relation network can be formed according to the user relation of the table. A relationship network is computationally infeasible. Therefore, according to the decomposition method of the relation network in the paper A sub-resolution triple center Algorithm for Large spark Networks with SmallMaximum Degreee, by Batagelj V, Mrvar A, the whole network is expressed by 13 subgraphs, and finally the number of the 13 subgraphs is taken as a vector to represent the whole network.
For example: we get such a user relationship in the first step:
| authors refer to | Relationship author | Relationships between |
| B | A | 1 |
| A | D | 2 |
| C | A | 1 |
Through the relationship, it can be known that the user B forwards the tweet of the user a, so that the user B and the tweet are in a forwarding relationship, and the user a specifies to forward the tweet of the user a to the user D, so that a specified relationship exists between the user B and the user D. User C forwards the tweet of user a so there is a forwarding relationship between the two. We can get the following relationship diagram according to the above relationship, as shown in fig. 1 a.
Since the user B forwards the tweet of the user a, the user B is a message obtained through the user a, and there is a directed line pointing to the user B by the user a, proving that the user B is a message obtained from the user a. Since user a specified that user D was forwarded to, user D is a message from user a, there is a directed line that user a points to user D. User C gets the message from user a because user C forwarded user a's tweet. There is a directed line that user a points to user C.
According to the decomposition method in the literature, the user relationship network can be subjected to subnet decomposition by grouping three points. User A, B, D constitutes sub-diagram 1 as in fig. 1b, user A, B, C constitutes sub-diagram 1 as in fig. 1c, and user A, B, D constitutes sub-diagram 1 as in fig. 1 d. There is no sub-graph 2-sub-graph 13 in the graph, so the graph will be expressed as a vector of [3,0,0,0,0,0,0, 0], with the vector indexed as sub-graph 1-sub-graph 13, with the data inside being the number of sub-graphs.
Step three: prediction
The change of a certain event in a certain period of the user relationship network is reflected on the change of the subgraphs, so that the period change of the number of the subgraphs is analyzed to predict whether the event has an outbreak trend in a period in the future.
Examples are: after the user relationship network of the first day of a certain event is decomposed and expressed, 14,0,0,555,0,0,0,121,0,0,0,0,0 are obtained, and it can be seen that the number of subgraphs 1 is 14, the number of subgraphs 4 is 555, the number of subgraphs 8 is 121, and the others are 0. After the user relationship network of the next day is decomposed and expressed, the number of sub-graphs 1 is 54, the number of sub-graphs 4 is 635, the number of sub-graphs 6 is 12, the number of sub-graphs 8 is 121, and the number of others is 0, so that the number of sub-graphs is 54,0,0,635,0, 121,0,0, 0. We can see that subgraph 1 and subgraph 4 grow rapidly in the two days, while subgraph 1 and subgraph 4 are mainly unilateral, which means that new users are added into the discussion all the time and the number of people is gradually enlarged, so we predict that the event still has a tendency to be enlarged on the third day.
Supplementing:
significance of thirteen subgraphs:
1. in sub-graph 1, sub-graph 2 and sub-graph 4, the user interaction is unidirectional, which indicates that the user group still expands outwards, and if the number of the three sub-graphs increases suddenly, the event is in a rapid expansion stage, and the known number of users increases suddenly.
2. In sub-diagram 3, sub-diagram 7 and sub-diagram 8, the user starts to have bidirectional interaction, which indicates that the discussion heat of the discussion of the user on a certain topic is increased, the interaction between the users is frequent, and if the number of the three sub-diagrams is increased, the topic heat is increased.
3. In sub-graph 5 and sub-graph 9, sub-graph closure begins to appear, and if the number of two sub-graphs increases, the sub-network of the local block is formed, the user growth is basically stable, and the topic popularity continues to increase.
4. In subgraph 6, subgraph 10, subgraph 11, subgraph 12 and subgraph 13, the subgraphs are closed and the bidirectional interaction is increased, and if five increases indicate that the number of users is basically stable, the topic heat is increased or the region is stable.
Claims (3)
1. A method for hot spot event outbreak prediction for twitter is characterized by comprising the following steps: the method comprises the following time steps:
step 1: obtaining data in twitter real-time data stream
Step 2: classification of tweets
Obtaining the tweet from the step 1, classifying the tweet by using the words in the text, and establishing a tweet classification data set;
and step 3: pushing context relationship combing
There are four cases for a tweet:
in the first case, the text is originated from the original, and there is no forwarding or forwarding to the specified object;
in the second case, the text is originated from the original, and there is a case of forwarding to a specific object;
in the third situation, the text is transferred from other authors, and the text is not transferred to the specified object;
the fourth case, text going from other authors, exists for forwarding to a particular object.
In the above four cases, the first case has no association relationship between users, and therefore, does not process it; in the second case, a piece of information is stored in the database, namely, the specified relationship existing between the author and the forwarded object is recorded; in the third case, recording the forwarding relation between the author and the forwarded tweet author; in the fourth case, two records are created, the forwarding relation between the author of the text and the author to be forwarded and the designated relation between the author of the text and the specific object to be forwarded are recorded respectively, and finally, a user relation data set of the tweet is formed;
and 4, step 4: prediction
The method makes a prediction for whether a topic can cause large-scale propagation, and the prediction step is divided into three steps:
the first step is as follows: organizing a user relationship network for a topic
The text pushing topic data set is obtained in the step 2, and for a certain topic, text pushing data of all the topics can be searched in the data set according to the ID of the topic; according to the ID of the tweet, obtaining the user relationship tables of all the tweets at present from the user relationship data set obtained in the step 3;
the second step is that: expressing a topic network
The first step obtains the user relation table, forms a user relation network according to the user relation of the table, and decomposes and expresses the user relation network according to the 13 seed graph model, namely [ l1,l2,…，l13]Wherein l isiDecomposing the representative user relationship network into the number of ith seed graph models;
step three: prediction
The change of a certain event in a certain period of the user relationship network is reflected on the change of the subgraphs, so that the period change of the number of the subgraphs is analyzed to predict whether the event has an outbreak trend in a period in the future.
2. The method of hotspot event burst prediction for twitter according to claim 1, wherein:
in step 1, the data comprises ID (tweet ID), user _ ID (tweet author ID), user _ location (tweet author country), profile (author head address), screen _ name (author nickname), content (content of the tweet), and created _ at (tweet creation time);
if the piece of tweet is forwarded from other authors, the following fields are also included in the data: retweeted (forwarded tweed content), retweeted _ ID (ID of forwarded tweed), retweeted _ screen _ name (nickname of forwarded tweed author), retweeted _ user _ ID (user ID of forwarded tweed), retweeted _ user _ location (country of forwarded tweed author), retweeted _ created _ at (creation time of forwarded tweed); each tweet contains a plurality of topics, and the relationship between the topics and the tweets is a many-to-many relationship.
3. The method of hotspot event burst prediction for twitter according to claim 1, wherein: the decomposition method of the user relationship network in the step 4 is a decomposition method mentioned by Batagelj V, Mrvar A in the paper "A SubqualaticTriad center Algorithm for Large spark Networks with Small Maximum Degreee", and the method relates to a 13-seed graph model for decomposing the network.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911390362.5A CN111143566A (en) | 2019-12-27 | 2019-12-27 | Method for predicting hot event outbreak aiming at twitter |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911390362.5A CN111143566A (en) | 2019-12-27 | 2019-12-27 | Method for predicting hot event outbreak aiming at twitter |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111143566A true CN111143566A (en) | 2020-05-12 |
Family
ID=70521677
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911390362.5A Pending CN111143566A (en) | 2019-12-27 | 2019-12-27 | Method for predicting hot event outbreak aiming at twitter |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111143566A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117556830A (en) * | 2024-01-11 | 2024-02-13 | 四川大学 | Rumor detection method based on potential hot topics and propagation process |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105809554A (en) * | 2016-02-07 | 2016-07-27 | 重庆邮电大学 | Prediction method of hot topics participated by users in social networks |
| US20170169018A1 (en) * | 2015-12-09 | 2017-06-15 | Le Holdings (Beijing) Co., Ltd. | Method and Electronic Device for Recommending Media Data |
| CN109829504A (en) * | 2019-02-14 | 2019-05-31 | 重庆邮电大学 | A kind of prediction technique and system forwarding behavior based on ICS-SVM analysis user |
| CN109857869A (en) * | 2019-01-26 | 2019-06-07 | 北京工业大学 | A kind of hot topic prediction technique based on Ap increment cluster and network primitive |
-
2019
- 2019-12-27 CN CN201911390362.5A patent/CN111143566A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170169018A1 (en) * | 2015-12-09 | 2017-06-15 | Le Holdings (Beijing) Co., Ltd. | Method and Electronic Device for Recommending Media Data |
| CN105809554A (en) * | 2016-02-07 | 2016-07-27 | 重庆邮电大学 | Prediction method of hot topics participated by users in social networks |
| CN109857869A (en) * | 2019-01-26 | 2019-06-07 | 北京工业大学 | A kind of hot topic prediction technique based on Ap increment cluster and network primitive |
| CN109829504A (en) * | 2019-02-14 | 2019-05-31 | 重庆邮电大学 | A kind of prediction technique and system forwarding behavior based on ICS-SVM analysis user |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117556830A (en) * | 2024-01-11 | 2024-02-13 | 四川大学 | Rumor detection method based on potential hot topics and propagation process |
| CN117556830B (en) * | 2024-01-11 | 2024-04-19 | 四川大学 | Rumor detection method based on potential hot topics and propagation process |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Bruns et al. | The use of Twitter hashtags in the formation of ad hoc publics | |
| Bindu et al. | Discovering spammer communities in twitter | |
| CN109063010B (en) | Opinion leader mining method based on PageRank | |
| Cheong et al. | Integrating web-based intelligence retrieval and decision-making from the twitter trends knowledge base | |
| Kwak et al. | What is Twitter, a social network or a news media? | |
| US8468144B2 (en) | Methods and apparatus for analyzing information to identify entities of significance | |
| EP3769278A1 (en) | Method of news evaluation in social media networks | |
| US9984109B2 (en) | Evolution aware clustering of streaming graphs | |
| US10366055B2 (en) | Decreasing duplicates and loops in an activity record | |
| Jin et al. | Research on social network structure and public opinions dissemination of micro-blog based on complex network analysis | |
| Pezzoni et al. | Why do I retweet it? An information propagation model for microblogs | |
| CN105608194A (en) | Method for analyzing main characteristics in social media | |
| Beskow et al. | You are known by your friends: Leveraging network metrics for bot detection in twitter | |
| Cotelo et al. | Dynamic topic‐related tweet retrieval | |
| Karagiannis et al. | Behavioral profiles for advanced email features | |
| Lim et al. | Tweets beget propinquity: Detecting highly interactive communities on twitter using tweeting links | |
| Abinaya et al. | Spam detection on social media platforms | |
| CN111143566A (en) | Method for predicting hot event outbreak aiming at twitter | |
| Zhang et al. | Information propagation in microblog networks | |
| Wolny | Knowledge gained from twitter data | |
| Agarwal et al. | Saturated total-population dependent branching process and viral markets | |
| CN112819645A (en) | Social network false information propagation detection method based on motif degree | |
| CN108989064B (en) | Topological data sampling method and device, visualization method and system | |
| Brandt et al. | Status and friendship: Mechanisms of social network evolution | |
| US11922345B2 (en) | Task management via a messaging service |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20200512 |