CN107451158B - Method for extracting semantic roles of traffic events in web text - Google Patents
Method for extracting semantic roles of traffic events in web text Download PDFInfo
- Publication number
- CN107451158B CN107451158B CN201610381179.9A CN201610381179A CN107451158B CN 107451158 B CN107451158 B CN 107451158B CN 201610381179 A CN201610381179 A CN 201610381179A CN 107451158 B CN107451158 B CN 107451158B
- Authority
- CN
- China
- Prior art keywords
- traffic
- information
- event
- event information
- role
- 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.)
- Expired - Fee Related
Links
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
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
-
- 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
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Human Computer Interaction (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Traffic Control Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a method for extracting traffic event semantic roles in a web text, which comprises the following steps of: the method comprises the steps of traffic event information structure definition, network text data preprocessing, traffic event information role labeling, event information role relationship pair extraction, traffic event role relationship tree construction, traffic positioning information element acquisition, traffic category information element acquisition, traffic event element relationship combination and traffic event information extraction. The invention provides a method for extracting semantic roles of traffic events in web texts by taking an open internet web text as a real-time data source of the traffic event information and utilizing the characteristics of high updating speed, more participators and wide user distribution of the web text, and provides effective supplement for the traditional traffic information acquisition means.
Description
Technical Field
The invention relates to a method for extracting traffic event semantic roles, in particular to a method for extracting traffic event semantic roles in a web text.
Background
Web text: the internet is a dynamic image of the whole society, multiple fields, wide depth and near real time, a huge internet information source is a main channel for the public to obtain information, and the quantity of internet web pages which can be indexed by 12 months in 2014 is 43.6 hundred million pages according to statistics.
Traffic events: the vehicle is in the event of personal injury or property loss caused by mistake or accident on the road. Not only by unspecified persons violating traffic regulations; or due to irresistible natural disasters such as earthquake, typhoon, mountain torrents, lightning stroke and the like.
Semantic role labeling: the method is a shallow semantic analysis implementation mode for natural language processing, and has the advantages of clear problem definition and convenience for manual marking and evaluation. The method does not perform detailed semantic analysis on the whole sentence, but only labels the semantic roles of some phrases in the sentence, for example, "committee will pass this protocol tomorrow" identifies in Chinese semantic role labeling corpus resource Chinese PropBank (CFB): the semantic role of "committee" is Arg0, which means the act of action; the semantic role of "this protocol" is labeled Arg1, which means that it represents the effect of an action; the semantic color notation of "tomorrow" is ArgM-TMP, which means time. The semantic roles Arg1, Arg0, ArgM-TMP above are all predefined semantic roles in CFB resources.
The traffic information acquisition technology is the core research content of an Intelligent Traffic System (ITS) and is also the basis for building the intelligent traffic system and carrying out traffic management. The traffic information is quickly, efficiently, accurately and comprehensively acquired, the overall operation efficiency of a traffic system can be improved, the occurrence frequency of traffic jam is reduced, the traffic safety is guaranteed, and the traffic management service level is improved. The traffic information mainly includes information such as road traffic flow, road conditions, traffic restrictions, traffic control, traffic events, construction work, road obstacles, traffic weather, road surface environment, and the like.
At present, the traffic information acquisition means mainly include the following four types:
1. the acquisition mode mainly depends on traffic information monitoring equipment which is deployed on a road in a large area in the early stage of urban traffic development, but the cost of the acquisition mode is too high, and the equipment replacement and maintenance are time-consuming and labor-consuming. In addition, the method can only acquire the section flow data of the road, the processing period of the data is slow, and the acquired overall traffic information is incomplete. At present, the method is mainly applied to monitoring important roads and key section layout of cities.
2. The floating car traffic information acquisition technology is the most mature technology in the aspect of traffic information acquisition at present and is the best information acquisition means for commercialization. The mode mainly depends on real-time position feedback of vehicles provided with positioning equipment (such as a GPS, a Beidou and the like), and the types of the vehicles mainly comprise taxies, buses, two-passenger and one-dangerous vehicles and the like. And the server side performs analysis processing operations such as map matching, speed extraction and the like on the returned position data in real time to acquire real-time traffic information of the road section. The advantage of this acquisition mode is that the traffic information coverage of gathering is higher on the one hand, can acquire the main urban road network in whole city, and on the other hand, the real-time is better, can in time know the traffic state in city, and in addition, compare with fixed sensor, equipment maintenance is simple, convenient, and the expense is lower.
3. The mobile phone signaling analysis traffic information acquisition technology can fully utilize mass mobile phone position resources of users, but has high requirements on data processing technology, and the mobile phone data acquisition has the risk of privacy disclosure. The technology has the advantages of all weather, wide coverage range and low cost, and can detect the real-time traffic information of suburban districts which can not be acquired by the floating car technology.
4. The technology for acquiring traffic information in the environment of the Internet of vehicles is characterized in that the Internet of vehicles is based on an in-vehicle network, an inter-vehicle network and a vehicle-mounted mobile internet, wireless communication and exchange of the vehicles and the internet are realized, and the acquisition of the traffic information can be realized by sensing and fusing related information of the vehicle network.
In summary, the four mainstream traffic information acquisition modes mainly acquire the road traffic state information, and besides the traffic state, the difficulty in acquiring other types of traffic information is still very high, and especially, the method has great limitations in the aspects of acquiring the sudden traffic information accurate to a point or the intersection traffic information, the temporary traffic control restriction information, dealing with the sudden traffic incident and the like,
at present, related research on traffic information acquisition technology mainly focuses on road condition information acquisition, and traffic red and green maps are the most well-known real-time road condition information. And relatively lack of research on information acquisition technologies such as traffic control, traffic accidents, traffic restrictions and the like. According to the statistics of the department of transportation and administration of public security, 238351 road traffic accidents occur nationwide in 2009; in 2010, road traffic accidents are reported in the whole country 3906164; in 2011, 210812 traffic accidents related to casualties occur all over the country, 68422 traffic accidents occur all over the golden period of eleven nations in 2012, and therefore, research on comprehensive traffic information acquisition technology is necessary; therefore, a method for extracting semantic roles of traffic events in web texts is needed, so that traffic event information can be extracted from the web texts, spatial position description of a traffic event linear reference method can be identified, and theme elements such as event occurrence spatial positions and events can be determined.
Disclosure of Invention
In order to solve the defects of the problems, the invention provides a method for extracting the semantic role of the traffic event in the web text.
In order to solve the above fixed problem, the solution adopted by the invention is as follows: a method for extracting traffic event semantic roles in web texts comprises the following extraction steps:
a. traffic event information structure definition:
the traffic incident information structure consists of incident information elements and incident information roles thereof;
the event information element comprises a positioning element and a category element;
the event information roles contained in the positioning elements are as follows: the road positioning system comprises an independent road, a non-independent road, an additional structure of the road, a positioning starting point, an additional structure of the positioning starting point, a positioning end point, an additional structure of the positioning end point, all directions, a starting direction and an ending direction;
the event information roles contained in the category elements are as follows: an event type;
b. preprocessing web text data:
carrying out data preprocessing on input network text data;
deleting repeated redundant information in the network text, and removing staying words;
chinese word segmentation of the web text to obtain word sequences after word segmentation;
c. and (3) traffic event information role labeling:
b, performing vocabulary event role labeling on the vocabulary sequence obtained in the step b by adopting a CRF conditional random field model to obtain a network text traffic event role sequence;
d. extracting event information role relationship pairs:
screening event information roles belonging to positioning elements from the event information role sequence, judging whether the screened event information roles are associated with each other by adopting an SVM (support vector machine) support vector machine model, and extracting all possible event information role relationship pairs;
e. constructing a traffic event role relationship tree:
d, aiming at the role relationship pair obtained in the step d, constructing a traffic event role relationship tree by adopting an algorithm RoleTreeBuild;
f. acquiring traffic positioning information elements:
extracting traffic incident information positioning elements from the traffic incident information role relationship tree obtained in the step e; adopting a depth-first traversal algorithm, starting from the vertex of the road where the path is located, searching a first unaccessed adjacent node, judging whether the traversal path contains the road where the path is located, positioning a starting point and a terminal point, and outputting the path if the traversal path contains the three elements;
g. acquiring traffic category information elements:
c, screening out event information roles belonging to category information elements from the event information role sequence obtained in the step c, and taking the event information roles as the category information elements of the traffic event information;
h. traffic event element relationship combination:
combining the category information element with the positioning information element obtained in the step d to form a positioning information element-category information element traffic information element pair;
i. extracting traffic event information:
and (e) aiming at the traffic information element pair set obtained in the step h, combining and extracting each information element pair to form traffic event information.
The extraction method further comprises the following steps: in the step c, a traffic incident information role labeling method adopting a CRF conditional random field model not only comprises the CRF conditional random field model, but also is suitable for the traffic incident information role labeling method protected by the patent;
and d, adopting an event information role relationship pair extraction method of an SVM (support vector machine) support vector machine model, wherein the method not only comprises an SVM conditional random field model, but also is suitable for the event information role relationship pair extraction method protected by the patent by using a hidden Markov model. The invention provides a method for extracting traffic event information in the network texts by taking the open Internet network texts as a real-time data source of the traffic event information and utilizing the characteristics of high updating speed, more participated people and wide user distribution, thereby providing effective supplement for the traditional traffic information acquisition means.
The method can be applied to online map websites, Location Based Services (LBS), travel navigation systems (such as Goods navigation and Baidu navigation) and the like, and provides complete traffic information data sources for the systems. In addition, the invention can also provide comprehensive traffic information support for the intelligent construction of the urban traffic system, thereby providing help for realizing more reasonable urban traffic planning, relieving traffic jam, reducing environmental pollution caused by traffic, improving the travel efficiency of people and leading traffic participants and traffic managers to know rich real-time traffic information in time.
Drawings
The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.
FIG. 1 is a flow chart of the extraction method of the present invention.
Fig. 2 is a schematic diagram of the construction of a traffic event role relationship tree according to the present invention.
Fig. 3 is a schematic diagram of acquiring a traffic event positioning information element according to the present invention.
Detailed Description
As shown in fig. 1, the flow chart of the extraction method of the present invention specifically includes the following steps:
1. traffic event information structure definition:
the traffic event information structure is composed of event information elements and event information roles thereof, and is defined as: wherein n is the number of traffic incident information elements, m is the number of semantic roles,
the event information element comprises a positioning element and a category element;
the event information roles contained in the positioning elements are as follows: the road positioning system comprises an independent road, a non-independent road, an additional structure of the road, a positioning starting point, an additional structure of the positioning starting point, a positioning end point, an additional structure of the positioning end point, all directions, a starting direction and an ending direction;
the event information roles contained in the category elements are as follows: an event type;
2. preprocessing web text data:
carrying out data preprocessing on input network text data: wherein p is the number of words;
deleting repeated redundant information in the network text, and removing staying words;
chinese word segmentation of the web text to obtain word sequences after word segmentation;
3. and (3) traffic event information role labeling:
labeling event roles with all vocabularies in the vocabulary sequence obtained in the step 2 by adopting a CRF conditional random field model, and obtaining a network text traffic event role sequence by referring to the role types;
4. extracting event information role relationship pairs:
screening event information roles belonging to positioning elements from the event information role sequence, judging whether the screened event information roles are associated with each other by adopting an SVM (support vector machine) support vector machine model, and extracting all possible event information role relationship pairs;
for example:
by training the SVM of the sample data set, we can obtain all possible role relationship pairs, for example, the combination of the road role rm and the positioning starting point role rs, and by training the data set, the combination of rm-rs is possible. It is possible to get all combinations of roles like rm-rs.
5. Constructing a traffic event role relationship tree:
and (4) constructing a traffic event role relation tree by adopting an algorithm RoleTreeBuild for the role relation pairs obtained in the step (4).
TABLE 1 traffic incident role relationship tree construction algorithm
6. Acquiring traffic positioning information elements:
extracting positioning elements of the traffic incident information from the traffic incident information role relationship tree obtained in the step 5; adopting a depth-first traversal algorithm, starting from the vertex of the road where the node is located, then finding out a first non-accessed adjacent node which just accesses the node, judging whether the traversal path contains three elements (the road where the node is located, a positioning starting point and a positioning end point), and if so, outputting the path;
7. acquiring traffic category information elements:
screening out event information roles belonging to category information elements from the event information role sequence obtained in the step 3, and taking the event information roles as the category information elements of the traffic event information;
8. traffic event element relationship combination:
combining the category information element and the positioning information element obtained in the step 6 into an information element pair with the type of 'positioning information element-category information element';
9. extracting traffic event information:
and (4) forming traffic event information by each information element pair in the information element pair set obtained in the step (8).
Example (b):
taking the sea as an example, the road sealing maintenance section for extracting the web text from "4 months, 21 days, later 24:00 to the next day, 5:00 comprises: the outer side of the inner ring is elevated and is provided with an outlet to the Nanpu bridge, and the inner side of the Nanpu bridge is provided with an inlet to the Cao-xi interchange; the south-north overhead west-side xu-Jia converging road reaches the Luban interchange entrance. "
1. Traffic event information structure definition:
the information structure of the traffic event information consists of event information elements and event roles under the information elements. The specific contents are as follows:
2. preprocessing web text data:
and 2, preprocessing the text, wherein the preprocessing comprises deleting repeated information in the web text and performing Chinese word segmentation on the web text to obtain a word sequence of the web text.
3. And (3) traffic event information role labeling:
and obtaining the role sequence of the network text events. As shown in the following table:
4. extracting the role relationship pair of the traffic incident:
the event information roles belonging to the positioning element are: the high-speed road comprises an inner ring elevated road (15), an outer side (16), a Longhua west road (17), an outlet (18), a Nanpu bridge (20), an inner side (22), a Nanpu bridge (23), a Cao-xi overpass (25), an inlet (26), a south-north elevated road (28), a west side (29), a Xujiahui road (30), a Luban road overpass (32) and an inlet (33).
The traffic event information role pair obtained by adopting the SVM support vector machine model comprises:
sm-rs | rm-fs | rm-sm |
rm-rs | rs-fs | rma-rm |
rm-rm | rm-fa | sm-rm |
rma-rma | sm-fa | rs-ss |
rs-re | rs-fa | re-se |
sm-fs | fs-fe | rs-se |
5. constructing a traffic event role relationship tree:
and (4) constructing a traffic event information role relationship tree by the role relationship pair obtained in the step 4, as shown in the following table and figure 2.
6. Acquiring a traffic event positioning information element:
the process flow, taking example web text as an example, is shown in fig. 3, and 2 sets of positioning information elements can be obtained:
on the road | Starting point of positioning | Positioning end point | |
|
Elevated outside of inner ring | Outlet of Longhua Xilu | Nanpu bridge |
|
Inner ring elevated inner side | Nanpu bridge | Cao xi grade separation entrance |
7. Acquiring traffic event category elements:
and (4) screening the event information roles belonging to the category information elements from the event information role sequence obtained in the step (3), and taking the event information roles as the category information elements of the traffic event information.
In the example web text, the event information roles belonging to the location information element are: and (6) sealing and maintaining (11). Thus, the category information elements derived from the example web text are: and (6) sealing and maintaining (11).
8. Traffic event element relationship combination:
and combining the category information element obtained in the step 7 and the positioning information element obtained in the step 6 into an information element pair with the type of positioning information element-category information element.
In an example web text, the information element pairs that are formed are:
|
|
|
|
9. Extracting traffic event information:
each information element pair in the set of information element pairs constitutes a traffic event information.
The traffic event information extracted from the example web text is:
those skilled in the art will appreciate that all or part of the steps in the method according to the above embodiments may be implemented by a program, which is stored in a computer-readable storage medium, and the program may be configured to: ROM/RAM, magnetic disk, optical disk, etc.
The above embodiments are not intended to limit the present invention, and the present invention is not limited to the above examples, and those skilled in the art may make variations, modifications, additions or substitutions within the technical scope of the present invention.
Claims (2)
1. A method for extracting traffic event semantic roles in web texts is characterized in that: the extraction steps are as follows:
a. traffic event information structure definition:
the traffic incident information structure consists of incident information elements and incident information roles thereof;
the event information element comprises a positioning element and a category element;
the event information roles contained in the positioning elements are as follows: the road positioning system comprises an independent road, a non-independent road, an additional structure of the road, a positioning starting point, an additional structure of the positioning starting point, a positioning end point, an additional structure of the positioning end point, all directions, a starting direction and an ending direction;
the event information roles contained in the category elements are as follows: an event type;
b. preprocessing web text data:
carrying out data preprocessing on input network text data;
deleting repeated redundant information in the network text, and removing staying words;
chinese word segmentation of the web text to obtain word sequences after word segmentation;
c. and (3) traffic event information role labeling:
b, performing vocabulary event role labeling on the vocabulary sequence obtained in the step b by adopting a CRF conditional random field model to obtain a network text traffic event role sequence;
d. extracting event information role relationship pairs:
screening event information roles belonging to positioning elements from the event information role sequence, judging whether the screened event information roles are associated with each other by adopting an SVM (support vector machine) support vector machine model, and extracting all possible event information role relationship pairs;
e. constructing a traffic event role relationship tree:
d, aiming at the role relationship pair obtained in the step d, constructing a traffic event role relationship tree by adopting an algorithm RoleTreeBuild;
f. acquiring traffic positioning information elements:
extracting traffic incident information positioning elements from the traffic incident information role relationship tree obtained in the step e; adopting a depth-first traversal algorithm, starting from the vertex of the road where the path is located, searching a first unaccessed adjacent node, judging whether the traversal path contains the road where the path is located, positioning a starting point and a terminal point, and outputting the path if the traversal path contains the three elements;
g. acquiring traffic category information elements:
c, screening out event information roles belonging to category information elements from the event information role sequence obtained in the step c, and taking the event information roles as the category information elements of the traffic event information;
h. traffic event element relationship combination:
combining the category information element with the positioning information element obtained in the step d to form a positioning information element-category information element traffic information element pair;
i. extracting traffic event information:
and (e) aiming at the traffic information element pair set obtained in the step h, combining and extracting each information element pair to form traffic event information.
2. The method for extracting semantic roles of traffic events from web texts according to claim 1, wherein the extraction method further comprises: in the step c, a traffic incident information role labeling method adopting a CRF conditional random field model not only comprises the CRF conditional random field model, but also is suitable for the traffic incident information role labeling method protected by the patent;
and d, adopting an event information role relationship pair extraction method of an SVM (support vector machine) support vector machine model, wherein the method not only comprises an SVM conditional random field model, but also is suitable for the event information role relationship pair extraction method protected by the patent by using a hidden Markov model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610381179.9A CN107451158B (en) | 2016-06-01 | 2016-06-01 | Method for extracting semantic roles of traffic events in web text |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610381179.9A CN107451158B (en) | 2016-06-01 | 2016-06-01 | Method for extracting semantic roles of traffic events in web text |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107451158A CN107451158A (en) | 2017-12-08 |
CN107451158B true CN107451158B (en) | 2021-01-19 |
Family
ID=60485256
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610381179.9A Expired - Fee Related CN107451158B (en) | 2016-06-01 | 2016-06-01 | Method for extracting semantic roles of traffic events in web text |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107451158B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110866177B (en) * | 2018-08-27 | 2022-07-01 | 中国石油化工股份有限公司 | Chemical accident news collecting method and machine readable storage medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102693314A (en) * | 2012-05-29 | 2012-09-26 | 代松 | Sensitive information monitoring method based on event search |
CN103268311A (en) * | 2012-11-07 | 2013-08-28 | 上海大学 | Event-structure-based Chinese statement analysis method |
CN103646088A (en) * | 2013-12-13 | 2014-03-19 | 合肥工业大学 | Product comment fine-grained emotional element extraction method based on CRFs and SVM |
CN103886080A (en) * | 2014-03-25 | 2014-06-25 | 中国科学院地理科学与资源研究所 | Method for extracting road traffic information from Internet unstructured text |
WO2014169857A1 (en) * | 2013-04-19 | 2014-10-23 | 富士通株式会社 | Data processing device, data processing method and electronic equipment |
CN105468666A (en) * | 2015-08-11 | 2016-04-06 | 中国科学院软件研究所 | Video content visual analysis method based on map metaphor |
-
2016
- 2016-06-01 CN CN201610381179.9A patent/CN107451158B/en not_active Expired - Fee Related
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102693314A (en) * | 2012-05-29 | 2012-09-26 | 代松 | Sensitive information monitoring method based on event search |
CN103268311A (en) * | 2012-11-07 | 2013-08-28 | 上海大学 | Event-structure-based Chinese statement analysis method |
WO2014169857A1 (en) * | 2013-04-19 | 2014-10-23 | 富士通株式会社 | Data processing device, data processing method and electronic equipment |
CN103646088A (en) * | 2013-12-13 | 2014-03-19 | 合肥工业大学 | Product comment fine-grained emotional element extraction method based on CRFs and SVM |
CN103886080A (en) * | 2014-03-25 | 2014-06-25 | 中国科学院地理科学与资源研究所 | Method for extracting road traffic information from Internet unstructured text |
CN105468666A (en) * | 2015-08-11 | 2016-04-06 | 中国科学院软件研究所 | Video content visual analysis method based on map metaphor |
Non-Patent Citations (3)
Title |
---|
"互联网文本蕴含道路交通信息抽取的模式匹配方法";仇培元 等;《地球信息科学学报》;20150430;第17卷(第4期);416-422 * |
"微博客蕴含交通事件信息抽取的自动标注方法";仇培元 等;《中文信息学报》;20170331;第31卷(第2期);107-116 * |
"领域本体学习中术语及关系抽取方法的研究";李丽双;《中国博士学位论文全文数据库 信息科技辑》;20130815(第8期);I138-194 * |
Also Published As
Publication number | Publication date |
---|---|
CN107451158A (en) | 2017-12-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Sakaki et al. | Real-time event extraction for driving information from social sensors | |
CN112133090A (en) | Multi-mode traffic distribution model construction method based on mobile phone signaling data | |
CN111653099B (en) | Bus passenger flow OD obtaining method based on mobile phone signaling data | |
CN106096631A (en) | A kind of recurrent population's Classification and Identification based on the big data of mobile phone analyze method | |
Liu et al. | Cluster analysis of microscopic spatio-temporal patterns of tourists’ movement behaviors in mountainous scenic areas using open GPS-trajectory data | |
CN106897919A (en) | With the foundation of car type prediction model, information providing method and device | |
CN110781670B (en) | Chinese place name semantic disambiguation method based on encyclopedic knowledge base and word vectors | |
CN105512166A (en) | Traffic parallel method with mapping between microblog public sentiments and city road conditions | |
CN101782897A (en) | Chinese corpus labeling method based on events | |
CN111931998B (en) | Individual travel mode prediction method and system based on mobile positioning data | |
Boyer et al. | Trends over two decades of transportation research: a machine learning approach | |
CN105893352A (en) | Air quality early-warning and monitoring analysis system based on big data of social network | |
CN111524353A (en) | Method for traffic text data for speed prediction and travel planning | |
Chen et al. | Rail transit ridership: station-area analysis of Boston’s Massachusetts Bay transportation authority | |
CN110597945B (en) | Cognitive site feature identification method and system for urban subway station domain | |
CN107451158B (en) | Method for extracting semantic roles of traffic events in web text | |
CN117455237A (en) | Road traffic accident risk prediction method based on multi-source data | |
CN113079463A (en) | Tourist attraction tourist travel activity identification method based on mobile phone signaling data | |
Qian | Real-time incident detection using social media data. | |
CN110610446A (en) | County town classification method based on two-step clustering thought | |
Fontes et al. | Leveraging social media as a source of mobility intelligence: An NLP-based approach | |
Wan et al. | Leveraging personal navigation assistant systems using automated social media traffic reporting | |
Ambastha et al. | Incident Detection From Social Media Targeting Indian Traffic Scenario Using Transfer Learning | |
Teng et al. | The method of analyzing metro complaint data and its application | |
Doran et al. | Discovering perceptions in online social media: A probabilistic approach |
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 | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20210119 |