CN111597351A - Visual document map construction method - Google Patents
Visual document map construction method Download PDFInfo
- Publication number
- CN111597351A CN111597351A CN202010404729.0A CN202010404729A CN111597351A CN 111597351 A CN111597351 A CN 111597351A CN 202010404729 A CN202010404729 A CN 202010404729A CN 111597351 A CN111597351 A CN 111597351A
- Authority
- CN
- China
- Prior art keywords
- node
- document
- document map
- construction method
- visual
- 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
- 230000000007 visual effect Effects 0.000 title claims abstract description 28
- 238000010276 construction Methods 0.000 title claims abstract description 20
- 238000000605 extraction Methods 0.000 claims abstract description 17
- 238000005065 mining Methods 0.000 claims abstract description 9
- 238000012800 visualization Methods 0.000 claims abstract description 7
- 238000004458 analytical method Methods 0.000 claims abstract description 5
- 230000011218 segmentation Effects 0.000 claims description 3
- 230000008520 organization Effects 0.000 claims description 2
- 238000000034 method Methods 0.000 abstract description 8
- 239000000284 extract Substances 0.000 abstract description 5
- 238000004364 calculation method Methods 0.000 description 8
- 230000007246 mechanism Effects 0.000 description 4
- 241000590419 Polygonia interrogationis Species 0.000 description 2
- 230000000295 complement effect Effects 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000003012 network analysis Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
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
- G06F40/205—Parsing
- G06F40/211—Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
-
- 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
- G06F16/367—Ontology
-
- 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
- G06F40/295—Named entity recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Abstract
The invention discloses a visual document map construction method, which comprises the following steps: s1, extracting keywords from the input text, segmenting each single sentence generated by segmenting the sentence, and sequentially performing part-of-speech tagging, named entity recognition and dependency syntactic analysis on each word formed by segmenting the words; s2: formulating a relation extraction rule, and extracting ternary group data from each single sentence obtained in S1 based on the relation extraction rule; the triple data consists of two entity words and a relation word; s3: importing the ternary group data obtained in the step S2 into a database to form a document map; s4: and carrying out graph mining operation on the data in the graph database, and realizing the visualization of the document graph based on the mining operation. The method and the device can extract the key information of the text, map the document into a visual graph based on semantic association, and help the user to efficiently master the semantic information of the article.
Description
Technical Field
The invention belongs to the technical field of knowledge maps, and particularly relates to a visual document map construction method.
Background
In the era of internet information explosion, people are difficult to quickly and accurately acquire the subject content of a document from unstructured text information, and especially the information overload is more serious for text documents such as reports, treatises, reports and the like. Therefore, it is necessary to perform a "dimension reduction" process on various types of texts. One common way is to form a text abstract, that is, to summarize and summarize an original text, and to summarize the main content of an article with a concise text, but this natural language information expression way does not allow people to intuitively and clearly obtain the required information. In 2012, google proposed a knowledge-graph concept. A knowledge graph is essentially a graph-based data structure, consisting of nodes (points) and edges (edges). Each node of the knowledge-graph represents an entity, and each edge is a relationship between the entities. Therefore, how to design a document map construction method based on the concept of the knowledge map to convert text information into a graphic format and simplify the content of graphics and texts to help users perceive and analyze article semantic information is a direction that needs to be researched by those skilled in the art.
Disclosure of Invention
The invention aims to provide a visual document map construction method, which can extract key information of a text, map the document into a visual map based on semantic association, simplify map complexity and help a user to efficiently master article information.
The technical scheme is as follows:
a visual document map construction method comprises the following steps: s1, extracting keywords from the input text, segmenting each single sentence generated by segmenting the sentence, and sequentially performing part-of-speech tagging, named entity recognition and dependency syntactic analysis on each word formed by segmenting the words; s2: formulating a relation extraction rule, and extracting ternary group data from each single sentence obtained in S1 based on the relation extraction rule; the triple data consists of two entity words and a relation word; s3: importing the ternary group data obtained in the step S2 into a database to form a document map; s4: and carrying out graph mining operation on the data in the graph database, and realizing the visualization of the document graph based on the mining operation.
By adopting the technical scheme: firstly, extracting key information from a document, mining semantic association among all entity words in an input text based on the extracted key information, and converting the semantic association into a node-edge-node map form. And the visualization of the atlas is implemented and optimized based on graph mining calculations. By mapping abstract data into graphic elements, a user is helped to effectively perceive and analyze article semantic information. In the above process, the text is divided into sentences based on commas, periods, exclamation marks and question marks. The named entity recognition process extracts terms of particular meaning, such as people, places, and organizations in the sentence.
Preferably, in the visual document map construction method: in step S1, keywords are extracted from the input text based on the TextRank algorithm.
By adopting the technical scheme: only nominal vocabularies in the text can be reserved to participate in the TextRank weight calculation, and a part of keywords can be filtered out by setting a threshold.
More preferably, in the visual document map construction method: introducing an external dictionary when carrying out named entity recognition in step S1; the external dictionary includes a person name, a place name, and an organization name.
By adopting the technical scheme: by introducing external dictionaries such as names, places, mechanisms and the like, the accuracy of named entity recognition can be effectively improved.
Further preferably, in the visual document map construction method: the relationship extraction rule in step S2 is constructed by combining dependency parsing with chinese grammar rules, which include isA rules and non-isA rules.
By adopting the technical scheme: in the dependency syntax structure, the main labeling relationships include: a cardinal relationship, a guest-moving relationship, a preposed object, a complement structure, a structure in a form, a guest-intervening relationship, etc. And mapping a plurality of structure combinations in the structures into a relation syntax rule, and applying the relation syntax rule to entity relation extraction. The isA rules are connected by the 'yes' family verb and comprise two types of rules of which the entity and the relation word have dependency relationship and rules of which the entity and the relation word have no dependency relationship. The non-isA rules are connected by verbs except the verb, and mainly comprise four types of syntax structures of a main predicate object, a main predicate intervening object, a main predicate supplement object and a preposed intervening object; in addition, the non-isA rules also include rules that the sentence has no subject, the rules indicate that the sentence has no subject, an entity can directly establish a moving guest relationship or indirectly establish a relationship such as a intervening guest and a moving guest, etc. with a verb, the previous sentence of the sentence is traced according to a Chinese heuristic rule, the subject of the previous sentence core verb is used as the subject of the sentence, and a corresponding triple is extracted.
Further preferably, in the visual document map construction method, the step S3 includes the following steps: s31: respectively establishing semantic-based association among the entity words; s32: respectively establishing association between the input text and each extracted entity word, storing the association in a graph database and giving a unique node name; s33: the document ID is given to each association obtained in S31 and S32.
By adopting the technical scheme: establishing semantic association refers to the fact that entity words in triples correspond to nodes in a graph database, relation words correspond to edges in the graph database, and the nodes are connected through the edges to form a relation network; the entity words are marked as different Label according to the types of the entity words; the sub-graph of each document may be obtained from the global graph by establishing associations of entities to the documents in S32. To avoid redundancy, the names of the nodes stored in the graph database are unique; by giving document ID (namely adding a fileID attribute) to each association, the affiliated documents of the relationship among the nodes are distinguished. The document ID adopts a unique identifier automatically generated by a system or a manually set unique identifier which is prestored.
Further preferably, in the visual document map construction method, the step S4 includes the following steps: s41: calculating the association number of each node and other nodes in the document map; s42: respectively calculating the weight of each node based on a node weight formula; s43: carrying out relationship combination on each node with the same relationship; s44: and importing the data obtained in the step S43 into a front end for visual display.
By adopting the technical scheme: the number of associations between each node and other nodes in the document graph obtained by the calculation in step S41 is the centrality of each node, which is the most direct measure for describing the centrality of the node in the network analysis. The more centralised a node is, the more cohesive the node is in the network. The graph shows the TOP N nodes spread out around the TOP with the highest centrality of the degree. Meanwhile, in step S42, the weight of each node is calculated, so as to set different node display sizes according to the weight, and highlight the emphasis of the text content. Through the steps of S41 and S42, the complexity of the finally generated graph network is simplified, and the understanding of the graph by the user is facilitated.
Still more preferably, in the visual document map construction method: the node weight formula in step S42 is
The W isijThe final weight value of the jth node of the ith document map is obtained; said DijThe association number of the jth node in the ith document map and other nodes in the document map is obtained; the T is the number of documents stored in a graph database; said N isjRepresenting the number of document maps containing node j.
By adopting the technical scheme: the weight of each node is calculated by the weight formula, so that the importance of different nodes in the same document can be highlighted, and the difference between documents containing certain same nodes can be highlighted.
Compared with the prior art, the method can extract the semantic relation among the entities in the document to form the visual document map based on the dependency syntactic relation extraction rule, can clearly and intuitively reveal the subject content of the article, highlights the key points and different points of the map content of each document, simplifies the complexity of the map network, and is convenient for a user to efficiently master the map information.
Drawings
The invention will be described in further detail with reference to the following detailed description and accompanying drawings:
FIG. 1 is a schematic view of the operation of example 1;
FIG. 2 is a schematic structural view in example 1;
FIG. 3 is a flowchart of extracting entity relationships in embodiment 1;
FIG. 4 is a visual document map formed after steps S1-S3;
fig. 5 is a visual document map finally formed after step S4.
Detailed Description
In order to more clearly illustrate the technical solution of the present invention, the following will be further described with reference to various embodiments.
Examples 1 are shown in FIGS. 1-3:
step S1, preprocessing the input text:
s11: extracting keywords of the input text based on a TextRank algorithm: only nominal vocabularies in the text are reserved to participate in the TextRank weight calculation, and then a threshold value is set and some key words are filtered;
the specific process is as follows: setting a threshold value as T, if the TextRank weight of the vocabulary is greater than the set threshold value T, keeping the words as key words, otherwise, filtering out the words
S12: sentence splitting is carried out on the input text: the text clause is aimed at analyzing the subsequent dependency syntax by taking a single sentence as a unit, and the adopted method is to divide the sentence according to punctuations such as commas, periods, exclamation marks, question marks and the like;
s13: performing word segmentation, part of speech tagging, named entity identification and dependency syntax processing on each single sentence: the named entity recognition extracts terms with specific meanings such as people, places and mechanisms in the sentence, and introduces external dictionaries such as names, places and mechanisms to improve the accuracy of the named entity recognition.
S21: establishing a relation extraction rule: and establishing a plurality of relation extraction rules by analyzing the dependency syntax relation and part-of-speech tagging among all the words in a single sentence.
S22: and (3) matching the relation rules: if an entity pair consisting of the keywords or the people, the places and the mechanisms obtained in the step S1 exists in the sentence, whether the entity pair meets a certain relation extraction rule is checked, and if the entity pair meets the certain relation extraction rule, a corresponding < entity 1, relation, entity 2> triple is output. And finally, after all sentences in the text are subjected to relation extraction, a text triple set can be obtained.
The relationship syntax rules are mainly constructed according to dependency syntax analysis and combined with heuristic rules of Chinese syntax, and a large number of entity relationship example analyses find that entity relationship triples always appear in certain fixed syntax structures, and verbs often play a connecting role in the structures, so that verbs connecting two entities can generally represent semantic relationships between the entities. In the dependency syntax structure, the main labeled relationships are shown in table 1 below, where the structure related to verbs is: a cardinal relationship, a guest-moving relationship, a preposed object, a complement structure, a structure in a form, a guest-intervening relationship, etc. Several structure combinations in the structures are mapped into a relation syntax rule, and the relation syntax rule can be applied to entity relation extraction. The invention divides the relation rules into two categories of isA rules and non-isA rules.
TABLE 1
The isA rules are connected by the "yes" family verb, and include two types of rules that the entity has dependency relationship with the relation word and the entity has no dependency relationship with the relation word, as shown in table 2 below:
TABLE 2
The non-isA rules are connected by verbs except the family verbs, and mainly include rules as shown in table 3 below:
TABLE 3
Further, the non-isA rules include rules without subject in the sentence in addition to the rules shown in table 3, the rules indicate that the subject does not exist in the sentence, an entity capable of directly establishing a moving object relationship with a verb or indirectly establishing an intervening object, a moving object, etc. relationship exists, the previous sentence in the sentence is traced according to the heuristic rule of chinese, the subject of the core verb of the previous sentence is used as the subject of the sentence, and the corresponding triples are extracted.
Step S3, importing the triple data extracted from the text into the database. For the data stored in the graph database, firstly, semantic association between entities is to be established, secondly, association between documents and extracted entity words is established, and the node names stored in the graph database are unique, so that homonymous nodes do not exist. The entity types and relationship types stored in the graph database are shown in Table 4 below:
TABLE 4
Further, to distinguish the document to which the relationship between the nodes belongs, a file ID (document ID) attribute is added to each relationship due to the uniqueness of the nodes.
At this time, the resulting map is shown in FIG. 4.
And step S4, carrying out graph mining calculation on the data in the graph database, and realizing the visualization of the document graph on the basis. The visualization mode of the graph is optimized through the series of graph calculation operations by calculating the degree centrality, the importance of the nodes and the same relation combination in the document graph. Specifically, the method comprises the following steps:
and calculating the centrality of the node degree in the graph database, selecting the Top N nodes with higher centrality in each document graph, and only displaying the associated networks of the N nodes. The degree centrality is used for calculating the relevance number of each node and other nodes to measure the cohesion of the nodes, and the nodes with higher degree centers are considered as key nodes in the network.
And calculating the weight of each node to distinguish the importance of the nodes in each document map, setting the size of the nodes according to the calculated node weight in map visualization, and highlighting the nodes with higher weight values when the document map is displayed. The idea of calculating the graph node weight refers to the TFIDF text feature extraction idea, and it is considered that the more frequently a node is associated with other nodes in a certain document map, and the less the node appears in other document maps, the more important the node is, that is, the larger the node weight value is. The node weight calculation formula is as follows:
wij represents the final weight value of the jth node of the ith document map; dij represents the correlation number between the jth node in the ith document map and other nodes in the document map, namely the degree centrality of the computing node; t represents the number of documents stored in a graph database, namely the number of nodes marked as 'File'; nj represents the number of document maps containing node j; for the same node Nj but different Dij in different document maps, the importance of different nodes in the same document can be highlighted through the weight calculation method, and the difference between documents containing certain same nodes can be highlighted.
And (5) combining the relations. If a node has a certain relationship with other nodes, the relationships are performed, for example, if there are triples < zhang, work 1>, < zhang, work 2>, the relationship is merged and becomes < zhang, work, [ work 1, work 2] >. The number of edges in the complex relationship network can be reduced after the maps are combined through the relationship so as to simplify the visual display of the maps. The resulting pattern is shown in FIG. 5.
The visual document map formed by the scheme can highlight the emphasis point of each document, can show the difference among different documents, can simplify the network complexity and is convenient for the map understanding of clients.
The above description is only an embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. The protection scope of the present invention is subject to the protection scope of the claims.
Claims (7)
1. A visual document map construction method is characterized by comprising the following steps:
s1: extracting keywords from the input text and segmenting the input text in sequence, segmenting each single sentence generated by the segmentation, and sequentially performing part-of-speech tagging, named entity recognition and dependency syntactic analysis on each word formed by segmentation;
s2: formulating a relation extraction rule, and extracting ternary group data from each single sentence obtained in S1 based on the relation extraction rule; the triple data consists of two entity words and a relation word;
s3: importing the ternary group data obtained in the step S2 into a database to form a document map;
s4: and carrying out graph mining operation on the data in the graph database, and realizing the visualization of the document graph based on the mining operation.
2. The visual document map construction method of claim 1, wherein: in step S1, keywords are extracted from the input text based on the TextRank algorithm.
3. The visual document map construction method of claim 2, wherein: introducing an external dictionary when carrying out named entity recognition in step S1; the external dictionary includes a person name, a place name, and an organization name.
4. The visual document map construction method of claim 3, wherein: the relationship extraction rule in step S2 is constructed by combining dependency parsing with chinese grammar rules, which include isA rules and non-isA rules.
5. The visual document map construction method according to claim 4, wherein the step S3 includes the steps of:
s31: respectively establishing semantic-based association among the entity words;
s32: respectively establishing association between the input text and each extracted entity word, storing the association in a graph database and giving a unique node name;
s33: the document ID is given to each association obtained in S31 and S32.
6. The visual document map construction method according to claim 5, wherein the step S4 includes the steps of:
s41: calculating the association number of each node and other nodes in the document map;
s42: respectively calculating the weight of each node based on a node weight formula;
s43: carrying out relationship combination on each node with the same relationship;
s44: and importing the data obtained in the step S43 into a front end for visual display.
7. The visual document map construction method of claim 6, wherein: the node weight formula in step S42 is:
the W isijThe final weight value of the jth node of the ith document map is obtained; said DijThe association number of the jth node in the ith document map and other nodes in the document map is obtained; the T is the number of documents stored in a graph database; said N isjRepresenting the number of document maps containing node j.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010404729.0A CN111597351A (en) | 2020-05-14 | 2020-05-14 | Visual document map construction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010404729.0A CN111597351A (en) | 2020-05-14 | 2020-05-14 | Visual document map construction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111597351A true CN111597351A (en) | 2020-08-28 |
Family
ID=72182243
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010404729.0A Pending CN111597351A (en) | 2020-05-14 | 2020-05-14 | Visual document map construction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111597351A (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112434532A (en) * | 2020-11-05 | 2021-03-02 | 西安交通大学 | Power grid environment model supporting man-machine bidirectional understanding and modeling method |
| CN112445915A (en) * | 2021-01-28 | 2021-03-05 | 京华信息科技股份有限公司 | Document map extraction method and device based on machine learning and storage medium |
| CN112463960A (en) * | 2020-10-30 | 2021-03-09 | 完美世界控股集团有限公司 | Entity relationship determination method and device, computing equipment and storage medium |
| CN112559760A (en) * | 2020-12-06 | 2021-03-26 | 复旦大学 | CPS (cyber physical system) resource capacity knowledge graph construction method for text description |
| CN112651226A (en) * | 2020-09-21 | 2021-04-13 | 深圳前海黑顿科技有限公司 | Knowledge analysis system and method based on dependency syntax tree |
| CN112671716A (en) * | 2020-12-03 | 2021-04-16 | 中国电子科技网络信息安全有限公司 | Vulnerability knowledge mining method and system based on map |
| CN112819160A (en) * | 2021-02-24 | 2021-05-18 | 文远鄂行(湖北)出行科技有限公司 | Visualization method, device, equipment and storage medium of neural network model |
| CN113312922A (en) * | 2021-04-14 | 2021-08-27 | 中国电子科技集团公司第二十八研究所 | Improved chapter-level triple information extraction method |
| CN113392183A (en) * | 2021-05-31 | 2021-09-14 | 南京师范大学 | Characterization and calculation method of children domain map knowledge |
| CN113486189A (en) * | 2021-06-08 | 2021-10-08 | 广州数说故事信息科技有限公司 | Open knowledge graph mining method and system |
| CN113609838A (en) * | 2021-07-14 | 2021-11-05 | 华东计算技术研究所（中国电子科技集团公司第三十二研究所） | Document information extraction and mapping method and system |
| CN113641786A (en) * | 2021-07-16 | 2021-11-12 | 中国人民解放军国防科技大学 | Document mixed recommendation method based on association graph and machine learning and application thereof |
| CN113705198A (en) * | 2021-10-21 | 2021-11-26 | 北京达佳互联信息技术有限公司 | Scene graph generation method and device, electronic equipment and storage medium |
| CN114040518A (en) * | 2021-11-26 | 2022-02-11 | 中国银行股份有限公司 | Network node display method and device |
| CN114997398A (en) * | 2022-03-09 | 2022-09-02 | 哈尔滨工业大学 | Knowledge base fusion method based on relation extraction |
| CN115409036A (en) * | 2022-08-15 | 2022-11-29 | 中国中医科学院中药研究所 | Dual-mode-based traditional Chinese medicine ancient book text entity relation extraction method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105786980A (en) * | 2016-02-14 | 2016-07-20 | 广州神马移动信息科技有限公司 | Method and apparatus for combining different examples for describing same entity and equipment |
| CN109241538A (en) * | 2018-09-26 | 2019-01-18 | 上海德拓信息技术股份有限公司 | Based on the interdependent Chinese entity relation extraction method of keyword and verb |
| CN110597999A (en) * | 2019-08-01 | 2019-12-20 | 湖北工业大学 | Judicial case knowledge graph construction method of dependency syntactic analysis relation extraction model |
-
2020
- 2020-05-14 CN CN202010404729.0A patent/CN111597351A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105786980A (en) * | 2016-02-14 | 2016-07-20 | 广州神马移动信息科技有限公司 | Method and apparatus for combining different examples for describing same entity and equipment |
| CN109241538A (en) * | 2018-09-26 | 2019-01-18 | 上海德拓信息技术股份有限公司 | Based on the interdependent Chinese entity relation extraction method of keyword and verb |
| CN110597999A (en) * | 2019-08-01 | 2019-12-20 | 湖北工业大学 | Judicial case knowledge graph construction method of dependency syntactic analysis relation extraction model |
Non-Patent Citations (1)
| Title |
|---|
| 朱鹏: "基于Web文档资源的课程知识图谱构建及应用研究", 《中国优秀硕士学位论文全文数据库社会科学Ⅱ辑》 * |
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112651226B (en) * | 2020-09-21 | 2022-03-29 | 深圳前海黑顿科技有限公司 | Knowledge analysis system and method based on dependency syntax tree |
| CN112651226A (en) * | 2020-09-21 | 2021-04-13 | 深圳前海黑顿科技有限公司 | Knowledge analysis system and method based on dependency syntax tree |
| CN112463960A (en) * | 2020-10-30 | 2021-03-09 | 完美世界控股集团有限公司 | Entity relationship determination method and device, computing equipment and storage medium |
| CN112463960B (en) * | 2020-10-30 | 2021-07-27 | 完美世界控股集团有限公司 | Entity relationship determination method and device, computing equipment and storage medium |
| CN112434532A (en) * | 2020-11-05 | 2021-03-02 | 西安交通大学 | Power grid environment model supporting man-machine bidirectional understanding and modeling method |
| CN112671716A (en) * | 2020-12-03 | 2021-04-16 | 中国电子科技网络信息安全有限公司 | Vulnerability knowledge mining method and system based on map |
| CN112671716B (en) * | 2020-12-03 | 2022-07-05 | 中国电子科技网络信息安全有限公司 | Vulnerability knowledge mining method and system based on map |
| CN112559760A (en) * | 2020-12-06 | 2021-03-26 | 复旦大学 | CPS (cyber physical system) resource capacity knowledge graph construction method for text description |
| CN112559760B (en) * | 2020-12-06 | 2022-04-12 | 复旦大学 | CPS (cyber physical system) resource capacity knowledge graph construction method for text description |
| CN112445915A (en) * | 2021-01-28 | 2021-03-05 | 京华信息科技股份有限公司 | Document map extraction method and device based on machine learning and storage medium |
| CN112819160A (en) * | 2021-02-24 | 2021-05-18 | 文远鄂行(湖北)出行科技有限公司 | Visualization method, device, equipment and storage medium of neural network model |
| CN112819160B (en) * | 2021-02-24 | 2023-10-31 | 文远鄂行(湖北)出行科技有限公司 | Visualization method, device and equipment for neural network model and storage medium |
| CN113312922A (en) * | 2021-04-14 | 2021-08-27 | 中国电子科技集团公司第二十八研究所 | Improved chapter-level triple information extraction method |
| CN113312922B (en) * | 2021-04-14 | 2023-10-24 | 中国电子科技集团公司第二十八研究所 | Improved chapter-level triple information extraction method |
| CN113392183A (en) * | 2021-05-31 | 2021-09-14 | 南京师范大学 | Characterization and calculation method of children domain map knowledge |
| CN113486189A (en) * | 2021-06-08 | 2021-10-08 | 广州数说故事信息科技有限公司 | Open knowledge graph mining method and system |
| CN113609838A (en) * | 2021-07-14 | 2021-11-05 | 华东计算技术研究所（中国电子科技集团公司第三十二研究所） | Document information extraction and mapping method and system |
| CN113641786B (en) * | 2021-07-16 | 2023-08-01 | 中国人民解放军国防科技大学 | Document mixed recommendation method based on association graph and machine learning and application thereof |
| CN113641786A (en) * | 2021-07-16 | 2021-11-12 | 中国人民解放军国防科技大学 | Document mixed recommendation method based on association graph and machine learning and application thereof |
| CN113705198A (en) * | 2021-10-21 | 2021-11-26 | 北京达佳互联信息技术有限公司 | Scene graph generation method and device, electronic equipment and storage medium |
| CN113705198B (en) * | 2021-10-21 | 2022-03-25 | 北京达佳互联信息技术有限公司 | Scene graph generation method and device, electronic equipment and storage medium |
| CN114040518A (en) * | 2021-11-26 | 2022-02-11 | 中国银行股份有限公司 | Network node display method and device |
| CN114997398A (en) * | 2022-03-09 | 2022-09-02 | 哈尔滨工业大学 | Knowledge base fusion method based on relation extraction |
| CN115409036A (en) * | 2022-08-15 | 2022-11-29 | 中国中医科学院中药研究所 | Dual-mode-based traditional Chinese medicine ancient book text entity relation extraction method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111597351A (en) | Visual document map construction method | |
| US20230142217A1 (en) | Model Training Method, Electronic Device, And Storage Medium | |
| CN104933027B (en) | A kind of open Chinese entity relation extraction method of utilization dependency analysis | |
| US8538979B1 (en) | Generating phrase candidates from text string entries | |
| Uma et al. | Formation of SQL from natural language query using NLP | |
| TWI656450B (en) | Method and system for extracting knowledge from Chinese corpus | |
| US20190171713A1 (en) | Semantic parsing method and apparatus | |
| CN111259160B (en) | Knowledge graph construction method, device, equipment and storage medium | |
| CN113220836A (en) | Training method and device of sequence labeling model, electronic equipment and storage medium | |
| CN114495143B (en) | Text object recognition method and device, electronic equipment and storage medium | |
| US20170357642A1 (en) | Cross Lingual Search using Multi-Language Ontology for Text Based Communication | |
| CN111984845B (en) | Website wrongly written word recognition method and system | |
| CN110929520A (en) | Non-named entity object extraction method and device, electronic equipment and storage medium | |
| CN111177401A (en) | Power grid free text knowledge extraction method | |
| CN114141384A (en) | Method, apparatus and medium for retrieving medical data | |
| WO2014002774A1 (en) | Synonym extraction system, method, and recording medium | |
| CN112527819A (en) | Address book information retrieval method and device, electronic equipment and storage medium | |
| CN112269884A (en) | Information extraction method, device, equipment and storage medium | |
| Mohnot et al. | Hybrid approach for Part of Speech Tagger for Hindi language | |
| US20230112385A1 (en) | Method of obtaining event information, electronic device, and storage medium | |
| CN115982322A (en) | Water conservancy industry design field knowledge graph retrieval method and retrieval system | |
| CN115510247A (en) | Method, device, equipment and storage medium for constructing electric carbon policy knowledge graph | |
| KR20220024251A (en) | Method and apparatus for building event library, electronic device, and computer-readable medium | |
| Jabbar et al. | A comparative review of Urdu stemmers: Approaches and challenges | |
| Parveen et al. | Clause Boundary Identification using Classifier and Clause Markers in Urdu Language |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200828 |