CN114186075A - Semantic search method for knowledge graph in culture field - Google Patents
Semantic search method for knowledge graph in culture field Download PDFInfo
- Publication number
- CN114186075A CN114186075A CN202111525616.7A CN202111525616A CN114186075A CN 114186075 A CN114186075 A CN 114186075A CN 202111525616 A CN202111525616 A CN 202111525616A CN 114186075 A CN114186075 A CN 114186075A
- Authority
- CN
- China
- Prior art keywords
- search
- result
- concept
- steps
- concepts
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 63
- 238000012163 sequencing technique Methods 0.000 claims abstract description 7
- 230000011218 segmentation Effects 0.000 claims description 26
- 238000004364 calculation method Methods 0.000 claims description 9
- 230000004927 fusion Effects 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 5
- 238000005516 engineering process Methods 0.000 claims description 4
- 238000009412 basement excavation Methods 0.000 claims description 3
- 238000004140 cleaning Methods 0.000 claims description 3
- 238000013135 deep learning Methods 0.000 claims description 3
- 238000007499 fusion processing Methods 0.000 claims description 3
- 230000008520 organization Effects 0.000 claims description 3
- 238000004806 packaging method and process Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000003672 processing method Methods 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
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- 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
- G06F16/3331—Query processing
- G06F16/334—Query execution
-
- 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 provides a semantic search method for knowledge maps in the culture field, which comprises a knowledge map body model in the culture field, wherein the body model is combined with search intentions to configure recommended associated terms, recommended term weight and result sequencing for search, and the knowledge map body model is fused with a full-text search engine for use based on search configuration.
Description
Technical Field
The invention belongs to the technical field of semantic search, and particularly relates to a semantic search method for knowledge maps in the culture field.
Background
In a cultural field knowledge content publishing site, a full-text search technology based on word segmentation is usually adopted for in-site search, the search principle is that after an article and a search sentence are subjected to word segmentation processing, word segmentation matching rate is calculated to obtain weight scores, then ranking is carried out according to the scores, the processing method cannot identify search intentions, if full-text fuzzy keyword matching search is carried out, contents having incidence relations with searched terms cannot be matched, the search method has great limitation, and the traditional word segmentation full-text search cannot understand the search intentions; the content related to the searched entry cannot be matched; the content of the incidence relation with the searched entry cannot be weighted, and therefore the invention provides a semantic searching method facing to the knowledge graph in the culture field.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides a semantic search method for knowledge base in the culture field, which is reasonable in design, adds an 'ancient historical character' model, sets character concept classification, simultaneously introduces character example information and character relations, adds character search configuration as the first three terms of relativity relations, the middle three terms of social relations and the last three terms of data as the written terms of users, and after storing configuration, uses the traditional search engine to search simultaneously with the method of the invention, the results of the traditional search engine search divides the names of people to match a large amount of characters and data containing the division and irrelevant to the searched content, and the search results of the method of the invention can accurately return the father, son and written works of the searched terms to be ordered in the front, after the keywords which are not related are matched with the result, the search experience and the search requirement can be effectively improved, and the search result is more accurate.
In order to achieve the purpose, the invention is realized by the following technical scheme: a semantic search method for knowledge maps in the culture field comprises a knowledge map body model in the culture field, wherein a recommended associated entry, recommended entry weight and result sequencing for searching are configured based on the body model and combined with search intentions, and the knowledge map body model is fused with a full-text search engine based on search configuration, and the semantic search method specifically comprises the following steps:
building a knowledge graph body model; the related cultural field metadata ontology model is adjusted and modified by the invention on the basis of a concept reference model (CIDOC-CRM) of the international literature working Committee, is suitable for constructing a knowledge map database of a conventional related subject knowledge base, and consists of an ontology concept classification part and a concept relation dictionary part; configuring a recommended associated entry, recommended entry weight and result sequencing of the search by combining the search intention based on the ontology model, wherein the method specifically comprises the following steps:
configuring a search query structure for different concepts using a search configuration system, the configuring steps comprising:
selecting a search intention concept: selecting concepts to be configured with search intentions from a configuration page, wherein the concepts are the attributions of preset search terms by concepts of intention identification hits;
secondly, the step of: adding a correlation concept; selecting concepts to be configured from a configuration page, wherein the concepts are concept attributions of recognized and hit preset search terms;
③: setting the number of results; setting the number of the associated concept query results;
fourthly, the method comprises the following steps: setting a result weight coefficient; the weight coefficient can participate in a formula to calculate the score value of each result, the score value influences the priority ranking of the associated concepts, the score calculation formula is equivalent to the score calculation of a full-text search engine, and the score calculation formula is as follows:
score(q，d)＝coord(q，d)×queryNorm(q)×∑t_in_q(tf(t_in_d)×idf(t)2×t.getBoost()×norm(t，d))
fifthly: defining an intention scene and storing; defining and storing the intention scene, and continuously adding the association concept to repeat the step if various association results exist after the intention scene is stored;
based on search configuration, the knowledge graph body model and a full-text search engine are fused for use, and the method specifically comprises the following steps:
processing the search; the method comprises the following specific steps:
the method comprises the following steps: searching by a user; a user submits a search request;
secondly, the step of: word segmentation; the method comprises the steps of performing natural semantic word segmentation on a search statement requested by a user, and extracting the part of speech of each word segmentation result, wherein the method adopts a Python-based Jieba plug-in to perform word segmentation and extract the part of speech, and in actual application, the word segmentation plug-in is not limited to only one plug-in;
③: named entity and intent recognition; combining word segmentation results, using a deep learning technology to identify named entities and search intentions for search sentences requested by a user by using a trained natural semantic recognition model to obtain concepts of the entities, adopting a lexical _ analysis pre-training model based on a PaddlePaddle framework to transform, expand and train, and carrying out Chinese lexical analysis and named entity recognition on the trained model results;
fourthly, the method comprises the following steps: acquiring a search configuration; acquiring configured concept search configuration according to the identified intention concept;
fifthly: searching a graph data path; obtaining associated entity results using a graph data path algorithm according to the obtained concept configuration;
sixthly, the method comprises the following steps: searching in full text; performing full-text search of text resources by using a full-text search engine according to a word segmentation result of a search sentence requested by a user;
seventh, the method comprises the following steps: fusing data; the method comprises the following steps of merging and fusing data of a path search result of a graph database and a full-text search result of a search engine, and sorting the weight reverse-order fusion result through the weight set by a configuration system, wherein the data fusion processing comprises the following substeps:
substep 1: the path result a of the graph data is prior to the result of the full text search b;
substep 2: after the results a are calculated and scored one by one, the results are arranged in a reverse order from large to small according to the scoring values;
substep 3: formatting the result a and the result b into a unified array data format;
substep 4: the array a is merged with the array b to obtain an array c (arr c [ ] [ arr a, arr b ]);
substep 5: carrying out duplicate removal cleaning on the array c;
and (v): outputting the result; and packaging the fusion result into a uniform interface and returning the format to the client.
In a preferred embodiment of the present invention, the named entity and the intention recognition, which are used by fusing the knowledge-graph body model with the full-text search engine based on the search configuration, are used in practical applications, and the named entity recognition method is not limited to only one method.
As a preferred embodiment of the present invention, the concept relationship dictionary of the cultural domain knowledge base ontology model is a basis for setting the relationship between the concept and the model, for example, "character" - "character" includes the relationships of father and son, mother and son, brother, sister, uncle and nephew, and "organization" - "movable cultural relic" includes the relationships of collection, excavation, circulation, etc.
The invention has the beneficial effects that:
the semantic search method facing the knowledge graph in the cultural field adds an ancient historical figure model, sets figure concept classification, simultaneously, introducing character example information and character relations, adding character search configuration that the first three are relatives relation entries, the middle three are social relation entries, the last three are written entries of the user, storing configuration, meanwhile, the traditional search engine and the method of the invention are used for searching, the name of the person is segmented according to the searching result of the traditional search engine, the results which contain the segmentation in a large amount of figure data but are irrelevant to the searched content are matched, the search result of the method can accurately return the father, son and bibliographic works of the search entry in the front and the keyword matching result with little relevance in the back, so that the search experience and the search requirement can be effectively improved, and the search result is more accurate.
Drawings
FIG. 1 is a conceptual diagram of an ontology model of a semantic search method for knowledge base in the culture field according to the present invention;
FIG. 2 is a schematic view of a search configuration flow of the semantic search method for knowledge base of cultural fields according to the present invention;
FIG. 3 is a schematic diagram of setting conditions of different intentions of character concepts in the semantic search method for knowledge maps in the cultural fields according to the present invention;
FIG. 4 is a schematic diagram of a search result scoring formula of the semantic search method for knowledge base in the culture field according to the present invention;
FIG. 5 is a schematic view of a search service flow of the semantic search method for knowledge base in culture field according to the present invention.
Detailed Description
In order to make the technical means, the creation characteristics, the achievement purposes and the effects of the invention easy to understand, the invention is further described with the specific embodiments.
Referring to fig. 1 to 5, the present invention provides a technical solution: a semantic search method for knowledge maps in the culture field comprises a knowledge map body model in the culture field, wherein a recommended associated entry, recommended entry weight and result sequencing for searching are configured based on the body model and combined with search intentions, and the knowledge map body model is fused with a full-text search engine based on search configuration, and the semantic search method specifically comprises the following steps:
building a knowledge graph body model; the related cultural field metadata ontology model is adjusted and modified by the invention on the basis of a concept reference model (CIDOC-CRM) of the international literature working Committee, is suitable for constructing a knowledge graph database of a conventional related subject knowledge base, and consists of an ontology concept classification part and a concept relation dictionary part, as shown in figure 1;
configuring a recommended associated entry, recommended entry weight and result sequencing of the search by combining the search intention based on the ontology model, wherein the method specifically comprises the following steps:
the search query structure is configured for different concepts using a search configuration system, as shown in FIG. 2, the configuration steps are as follows:
selecting a search intention concept: selecting concepts to be configured with search intentions from a configuration page, wherein the concepts are the attributions of preset search terms by concepts of intention identification hits;
secondly, the step of: adding a correlation concept; selecting concepts to be configured from a configuration page, wherein the concepts are the concept attributions of the recognized and hit preset search terms, as shown in FIG. 3;
③: setting the number of results; setting the number of the associated concept query results;
fourthly, the method comprises the following steps: setting a result weight coefficient; the weight coefficient can participate in a formula to calculate the score value of each result, the score value influences the priority ranking of the associated concepts, the score calculation formula is equivalent to the score calculation of a full-text search engine, and the score calculation formula is as follows:
score(q，d)＝coord(q，d)×queryNorm(q)×∑t_in_q(tf(t_in_d)×idf(t)2×t.getBoost()×norm(t，d))
fifthly: defining an intention scene and storing; defining and storing the intention scene, and continuously adding the association concept to repeat the step if various association results exist after the intention scene is stored;
based on search configuration, the knowledge graph body model and a full-text search engine are fused for use, and the method specifically comprises the following steps:
processing the search; as shown in fig. 5, the specific steps are as follows:
the method comprises the following steps: searching by a user; a user submits a search request;
secondly, the step of: word segmentation; the method comprises the steps of performing natural semantic word segmentation on a search statement requested by a user, and extracting the part of speech of each word segmentation result, wherein the method adopts a Python-based Jieba plug-in to perform word segmentation and extract the part of speech, and in actual application, the word segmentation plug-in is not limited to only one plug-in;
③: named entity and intent recognition; combining word segmentation results, using a deep learning technology to identify named entities and search intentions for search sentences requested by a user by using a trained natural semantic recognition model to obtain concepts of the entities, adopting a lexical _ analysis pre-training model based on a PaddlePaddle framework to transform, expand and train, and carrying out Chinese lexical analysis and named entity recognition on the trained model results;
fourthly, the method comprises the following steps: acquiring a search configuration; acquiring configured concept search configuration according to the identified intention concept;
fifthly: searching a graph data path; obtaining associated entity results using a graph data path algorithm according to the obtained concept configuration;
sixthly, the method comprises the following steps: searching in full text; performing full-text search of text resources by using a full-text search engine according to a word segmentation result of a search sentence requested by a user;
seventh, the method comprises the following steps: fusing data; the method comprises the following steps of merging and fusing data of a path search result of a graph database and a full-text search result of a search engine, and sorting the weight reverse-order fusion result through the weight set by a configuration system, wherein the data fusion processing comprises the following substeps:
substep 1: the path result a of the graph data is prior to the result of the full text search b;
substep 2: after the results a are calculated and scored one by one, the results are arranged in a reverse order from large to small according to the scoring values;
substep 3: formatting the result a and the result b into a unified array data format;
substep 4: the array a is merged with the array b to obtain an array c (arr c [ ] [ arr a, arr b ]);
substep 5: carrying out duplicate removal cleaning on the array c;
and (v): outputting the result; and packaging the fusion result into a uniform interface and returning the format to the client.
In a preferred embodiment of the present invention, the named entity and the intention recognition, which are used by fusing the knowledge-graph body model with the full-text search engine based on the search configuration, are used in practical applications, and the named entity recognition method is not limited to only one method.
As a preferred embodiment of the present invention, the concept relationship dictionary of the cultural domain knowledge base ontology model is a basis for setting the relationship between the concept and the model, for example, "character" - "character" includes the relationships of father and son, mother and son, brother, sister, uncle and nephew, and "organization" - "movable cultural relic" includes the relationships of collection, excavation, circulation, etc.
While there have been shown and described what are at present considered the fundamental principles and essential features of the invention and its advantages, it will be apparent to those skilled in the art that the invention is not limited to the details of the foregoing exemplary embodiments, but is capable of other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Furthermore, it should be understood that although the present description refers to embodiments, not every embodiment may contain only a single embodiment, and such description is for clarity only, and those skilled in the art should integrate the description, and the embodiments may be combined as appropriate to form other embodiments understood by those skilled in the art.
Claims (3)
1. A semantic search method for knowledge maps in the culture field is characterized by comprising a knowledge map body model in the culture field, recommending associated entries, recommending entry weights and result sequencing which are searched based on the body model and combined with search intention configuration, and fusing the knowledge map body model with a full-text search engine based on search configuration for use, wherein the semantic search method specifically comprises the following steps:
building a knowledge graph body model; the related cultural field metadata ontology model is adjusted and modified by the invention on the basis of a concept reference model (CIDOC-CRM) of the international literature working Committee, is suitable for constructing a knowledge map database of a conventional related subject knowledge base, and consists of an ontology concept classification part and a concept relation dictionary part;
configuring a recommended associated entry, recommended entry weight and result sequencing of the search by combining the search intention based on the ontology model, wherein the method specifically comprises the following steps:
configuring a search query structure for different concepts using a search configuration system, the configuring steps comprising:
selecting a search intention concept: selecting concepts to be configured with search intentions from a configuration page, wherein the concepts are the attributions of preset search terms by concepts of intention identification hits;
secondly, the step of: adding a correlation concept; selecting concepts to be configured from a configuration page, wherein the concepts are concept attributions of recognized and hit preset search terms;
③: setting the number of results; setting the number of the associated concept query results;
fourthly, the method comprises the following steps: setting a result weight coefficient; the weight coefficient can participate in a formula to calculate the score value of each result, the score value influences the priority ranking of the associated concepts, the score calculation formula is equivalent to the score calculation of a full-text search engine, and the score calculation formula is as follows:
score(q，d)＝coord(q，d)×queryNorm(q)×∑t_in_q(tf(t_in_d)×idf(t)2×t.getBoost()×norm(t，d))
fifthly: defining an intention scene and storing; defining and storing the intention scene, and continuously adding the association concept to repeat the step if various association results exist after the intention scene is stored;
based on search configuration, the knowledge graph body model and a full-text search engine are fused for use, and the method specifically comprises the following steps:
processing the search; the method comprises the following specific steps:
the method comprises the following steps: searching by a user; a user submits a search request;
secondly, the step of: word segmentation; the method comprises the steps of performing natural semantic word segmentation on a search statement requested by a user, and extracting the part of speech of each word segmentation result, wherein the method adopts a Python-based Jieba plug-in to perform word segmentation and extract the part of speech, and in actual application, the word segmentation plug-in is not limited to only one plug-in;
③: named entity and intent recognition; combining word segmentation results, using a deep learning technology to identify named entities and search intentions for search sentences requested by a user by using a trained natural semantic recognition model to obtain concepts of the entities, adopting a lexical _ ahalysis pre-training model based on a PaddlePaddle framework to modify, expand and train, and carrying out Chinese lexical analysis and named entity recognition on the trained model results;
fourthly, the method comprises the following steps: acquiring a search configuration; acquiring configured concept search configuration according to the identified intention concept;
fifthly: searching a graph data path; obtaining associated entity results using a graph data path algorithm according to the obtained concept configuration;
sixthly, the method comprises the following steps: searching in full text; performing full-text search of text resources by using a full-text search engine according to a word segmentation result of a search sentence requested by a user;
seventh, the method comprises the following steps: fusing data; the method comprises the following steps of merging and fusing data of a path search result of a graph database and a full-text search result of a search engine, and sorting the weight reverse-order fusion result through the weight set by a configuration system, wherein the data fusion processing comprises the following substeps:
substep 1: the path result a of the graph data is prior to the result of the full text search b;
substep 2: after the results a are calculated and scored one by one, the results are arranged in a reverse order from large to small according to the scoring values;
substep 3: formatting the result a and the result b into a unified array data format;
substep 4: the array a is merged with the array b to obtain an array c (arr c [ ] [ arr a, arr b ]);
substep 5: carrying out duplicate removal cleaning on the array c;
and (v): outputting the result; and packaging the fusion result into a uniform interface and returning the format to the client.
2. The cultural domain knowledge graph-oriented semantic search method according to claim 1, wherein: the named entity and intention recognition based on search configuration and fusing the knowledge map body model and the full-text search engine is applied to practical application, and the named entity recognition method is not limited to only one method.
3. The cultural domain knowledge graph-oriented semantic search method according to claim 1, wherein: the concept relation dictionary of the cultural field knowledge base ontology model is a basis for setting the relation between model concepts and concepts, such as 'character' -character 'comprising the relations of father and son, mother and son, brother, sister, uncle and nephew and the like, and' organization '-movable cultural relic' comprising the relations of collection, excavation, stream and the like.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111525616.7A CN114186075B (en) | 2021-12-14 | 2021-12-14 | Semantic search method for knowledge graph in cultural domain |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111525616.7A CN114186075B (en) | 2021-12-14 | 2021-12-14 | Semantic search method for knowledge graph in cultural domain |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114186075A true CN114186075A (en) | 2022-03-15 |
| CN114186075B CN114186075B (en) | 2024-04-26 |
Family
ID=80543739
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111525616.7A Active CN114186075B (en) | 2021-12-14 | 2021-12-14 | Semantic search method for knowledge graph in cultural domain |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114186075B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170024375A1 (en) * | 2015-07-26 | 2017-01-26 | Microsoft Technology Licensing, Llc | Personal knowledge graph population from declarative user utterances |
| CN109522465A (en) * | 2018-10-22 | 2019-03-26 | 国家电网公司 | The semantic searching method and device of knowledge based map |
| JP2019074843A (en) * | 2017-10-13 | 2019-05-16 | ヤフー株式会社 | Information providing apparatus, information providing method, and program |
| CN110704743A (en) * | 2019-09-30 | 2020-01-17 | 北京科技大学 | Semantic search method and device based on knowledge graph |
| CN113190593A (en) * | 2021-05-12 | 2021-07-30 | 《中国学术期刊（光盘版）》电子杂志社有限公司 | Search recommendation method based on digital human knowledge graph |
-
2021
- 2021-12-14 CN CN202111525616.7A patent/CN114186075B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170024375A1 (en) * | 2015-07-26 | 2017-01-26 | Microsoft Technology Licensing, Llc | Personal knowledge graph population from declarative user utterances |
| JP2019074843A (en) * | 2017-10-13 | 2019-05-16 | ヤフー株式会社 | Information providing apparatus, information providing method, and program |
| CN109522465A (en) * | 2018-10-22 | 2019-03-26 | 国家电网公司 | The semantic searching method and device of knowledge based map |
| CN110704743A (en) * | 2019-09-30 | 2020-01-17 | 北京科技大学 | Semantic search method and device based on knowledge graph |
| CN113190593A (en) * | 2021-05-12 | 2021-07-30 | 《中国学术期刊（光盘版）》电子杂志社有限公司 | Search recommendation method based on digital human knowledge graph |
Non-Patent Citations (1)
| Title |
|---|
| 郭嘉欣;: "基于多源异构数据挖掘的"红色记忆"知识图谱构建", 知识管理论坛, no. 01, 28 February 2020 (2020-02-28) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114186075B (en) | 2024-04-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106997382B (en) | Innovative creative tag automatic labeling method and system based on big data | |
| CN107122413A (en) | A kind of keyword extracting method and device based on graph model | |
| CN102637192A (en) | Method for answering with natural language | |
| CN110020189A (en) | A kind of article recommended method based on Chinese Similarity measures | |
| CN103744984B (en) | Method of retrieving documents by semantic information | |
| CN109960756A (en) | Media event information inductive method | |
| CN110888991A (en) | Sectional semantic annotation method in weak annotation environment | |
| CN112036178A (en) | Distribution network entity related semantic search method | |
| CN108984521A (en) | Personage's viewpoint abstracting method in a kind of media event | |
| CN112749265A (en) | Intelligent question-answering system based on multiple information sources | |
| Bellare et al. | Lightly-supervised attribute extraction | |
| CN114064901B (en) | Book comment text classification method based on knowledge graph word meaning disambiguation | |
| CN114841353A (en) | Quantum language model modeling system fusing syntactic information and application thereof | |
| CN112084312B (en) | Intelligent customer service system constructed based on knowledge graph | |
| CN112417170B (en) | Relationship linking method for incomplete knowledge graph | |
| Al-Sultany et al. | Enriching tweets for topic modeling via linking to the wikipedia | |
| CN114186075A (en) | Semantic search method for knowledge graph in culture field | |
| CN112989811B (en) | History book reading auxiliary system based on BiLSTM-CRF and control method thereof | |
| CN112445887B (en) | Method and device for realizing machine reading understanding system based on retrieval | |
| Doostmohammadi et al. | Perkey: A persian news corpus for keyphrase extraction and generation | |
| CN114661856A (en) | Fusion map construction method | |
| Han et al. | Construction method of knowledge graph under machine learning | |
| CN111858885A (en) | Keyword separation user question intention identification method | |
| Lin et al. | Domain Independent Key Term Extraction from Spoken Content Based on Context and Term Location Information in the Utterances | |
| CN112948544B (en) | Book retrieval method based on deep learning and quality influence |
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 |