CN113204648A - Knowledge graph completion method based on automatic extraction relationship of judgment book text - Google Patents
Knowledge graph completion method based on automatic extraction relationship of judgment book text Download PDFInfo
- Publication number
- CN113204648A CN113204648A CN202110478354.7A CN202110478354A CN113204648A CN 113204648 A CN113204648 A CN 113204648A CN 202110478354 A CN202110478354 A CN 202110478354A CN 113204648 A CN113204648 A CN 113204648A
- Authority
- CN
- China
- Prior art keywords
- text
- predicate
- knowledge
- knowledge graph
- decision
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000000605 extraction Methods 0.000 title claims abstract description 22
- 238000004590 computer program Methods 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 4
- 230000000295 complement effect Effects 0.000 abstract description 3
- 238000010276 construction Methods 0.000 abstract description 2
- 230000006870 function Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000013519 translation 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
- G06N5/022—Knowledge engineering; Knowledge acquisition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/18—Legal services; Handling legal documents
Abstract
The invention provides a knowledge graph complementing method for automatically extracting relations based on a judgment book text, which comprises the steps of setting predicate guide words by summarizing the existing construction method of the knowledge graph and extracting the relations in a specific field; summarizing and warehousing the relationships which appear in the new judgment book text and cannot be extracted by the existing method so as to guide the existing knowledge graph to add new relationships to complement the knowledge graph; the above steps are repeated in a circulating way and dynamically updated to form a continuously updated and robust closed loop flow, and the functions of automatically extracting triples in the text of the judgment book and dynamically improving and enriching the content of the knowledge graph are realized. The relationship extraction method for completion is established on the knowledge graph, so that the classification is more accurate and the coverage is more comprehensive. The invention completes different knowledge maps according to the judgment book texts of different cases, completes specific completion tasks aiming at different types of judgment books, and has stronger pertinence and practicability.
Description
Technical Field
The invention belongs to the technical field of knowledge graph completion, and particularly relates to a knowledge graph completion method based on automatic extraction relation of judgment book texts.
Background
At present, the application of knowledge maps in various industries is gradually opened, particularly in the aspects of finance, medical treatment, law and tourism. Knowledge map assistance legal intelligence can utilize current big data and natural language processing technology to a certain extent, provides some intelligent solutions. Due to the information updating speed, the dynamic knowledge graph completion can break through a fixed scene, process new entities or new relations, and have strong expansibility. The legal knowledge map has better practical significance. So it is necessary to complement knowledge-graph based on automatic extraction.
Knowledge Graph (Knowledge Graph) is a structured semantic Knowledge base, which describes concept entities and their interrelations in the physical world in the form of symbols, and analyzes problems from the perspective of "relations". Therefore, the relation extraction is a key step of knowledge graph completion. The information source of knowledge graph completion mainly comprises two aspects: firstly, discovering new relations existing in two old entities; second, to find the relationship between two new entities. The knowledge graph has certain significance no matter what kind of information source is complemented so as to make the knowledge graph dynamic. The knowledge graph based on the decision book is characterized by strong domain and normalization and small granularity. And the supplement method based on the legal knowledge map is very few. The traditional completion algorithm is a knowledge graph completion algorithm based on translation transformation and using Euclidean distance as measurement, the accuracy of the model is interfered by irrelevant dimensions, and the model has limitation when complex relations are processed; the method is applied to completion of the decision book knowledge graph and is a small topic. Secondly, a knowledge graph completion algorithm based on the relation path cannot effectively calculate on a low-connectivity graph (a knowledge graph with sparse relation); if the method is adopted, a mature and perfect judgment book knowledge graph is required to be used as a basis.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the method is used for dynamically and automatically completing the knowledge graph.
The technical scheme adopted by the invention for solving the technical problems is as follows: a knowledge graph completion method for automatically extracting relations based on a decision book text comprises the following steps:
s01: constructing a legal knowledge graph according to the types of the judgment book texts;
s02: extracting the relation in the legal knowledge graph;
s06: extracting the relation of sentences in the text of the decision book according to the relation obtained in the step S02, and executing a step S07 if the extraction is successful; if the extraction is not possible, go to step S08;
s07: adding the relationship extracted in step S06 to the legal knowledge base;
s08: the relationship that could not be extracted in step S06 is added to the predicate-oriented thesaurus and the legal knowledge graph is complemented.
According to the scheme, in the step S02, the specific steps are as follows: and extracting predicate verbs of the triples in the legal knowledge graph and setting a predicate-oriented lexicon.
Further, in step S06, the specific steps include: the predicate-directors in the predicate-director lexicon are used to match the sentences in the decision-book text one by one.
Further, in step S07, the specific steps include: and establishing a triple through the sentences matched with the predicate guidance words, and adding the triple into the legal knowledge graph.
Further, in step S08, the specific steps include: and screening sentences which are not matched with the predicate guide words, adding new predicate verbs into a predicate guide word library, and completing the legal knowledge graph.
According to the scheme, the method between the step S02 and the step S06 further comprises the following steps:
s03: selecting different legal knowledge maps and predicate guide lexicons according to the types of the judgment book texts;
s04: setting a starting position and an ending position of a judgment book text;
s05: in the text of the judgment bookThe sentences between the starting position and the ending position are numbered as a in sequence1、…、anAnd parts of speech are labeled to the components of the sentence.
Further, in step S08, sentences that do not match the predicate director are sequentially numbered as b1、…、bn(ii) a Sequentially numbering new predicate directors added into the predicate-oriented lexicon as C1、…、Cn。
According to the scheme, the method further comprises the following steps:
s09: and selecting the judgment text for constructing the legal knowledge graph, and executing the steps S04 to S06 on the existing judgment text by using the new predicate guide words.
According to the scheme, the method further comprises the following steps:
s10: executing the step S03 to the step S09 to the newly added judgment book text;
s11: and completing the legal knowledge map according to the newly added judgment book text.
A computer storage medium having stored therein a computer program executable by a computer processor, the computer program performing a method of knowledgegraph replenishment of automatically extracted relationships based on decision book text.
The invention has the beneficial effects that:
1. the invention relates to a knowledge graph complementing method based on automatic extraction relation of a decision book text, which is a circular and gradual process integrating manual participation and automatic extraction; on one hand, the construction method of the existing knowledge graph including the legal field is summarized through manual participation, such as triples, predicate guidance words are set, and the relation of the specific field including the new judgment text is extracted; on the other hand, the relationships which appear in the new judgment book text and cannot be extracted by the existing method are summarized and put in storage so as to guide the existing knowledge graph to add new relationships to complement the knowledge graph; the above steps are repeated in a circulating way and dynamically updated to form a continuously updated and robust closed loop flow, and the functions of automatically extracting triples in the text of the judgment book and dynamically improving and enriching the content of the knowledge graph are realized.
2. The relation extraction method for completion is established on the knowledge graph, and the extracted range is more specific and accurate by setting the predicate guide words by summarizing triples of the knowledge graph, so that the established triples have higher utilization rate for completion of the knowledge graph; therefore, the invention has more accurate classification and more comprehensive coverage.
3. The invention completes different knowledge maps according to the judgment book texts of different cases, completes specific completion tasks aiming at different types of judgment books, and has stronger pertinence and practicability.
Drawings
FIG. 1 is a flow chart of an embodiment of the present invention.
FIG. 2 is a pseudo-card swipe domain knowledge graph of an embodiment of the invention.
FIG. 3 is an exemplary diagram of sentence numbers and part-of-speech tags according to an embodiment of the present invention.
Fig. 4 is a schematic diagram of the fake card embezzlement completion according to the embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.
Referring to fig. 1, a knowledge graph completion method for automatically extracting relationships based on a decision book text according to an embodiment of the present invention includes the following steps:
s01: the domain expert constructs different legal knowledge maps according to the types of the judgment book texts;
s02: setting a predicate-oriented lexicon by extracting predicate verbs of triples in the legal knowledge graph;
s03: selecting different knowledge maps and predicate-oriented word banks according to the types of the judgment books;
s04: setting the starting and ending positions of the text of the judgment book (such as the starting position: the court trial is found out, and the ending position: the court deems);
s05: numbering sentences in starting and stopping positions in legal decision books (a)1…an) And performing part-of-speech tagging;
s06: matching the predicate-oriented words in the selected predicate-oriented lexicon one by one with the new legal decision textSentence (a)1…an)；
S07: establishing a triple of the sentences matched with the predicate guidance words and adding the triple into the existing knowledge graph;
s08: labeling sentences not matched with predicate director (b)1…bn)；
S09: sentence b of reference number1…bnScreening is carried out, new predicate verbs are summarized, whether the new predicate verbs are added into a predicate-oriented word library or not is judged, and if new predicate-oriented words are stored, the new predicate-oriented words are labeled (C)1…Cn) Selecting a legal decision book text for constructing a knowledge graph;
s10: continuing steps S4-S6 for the existing legal adjudication with the new predicate guide;
s11: repeating steps S3-S10 whenever a new addition of legal decision text occurs;
s12: and completing the completion work of the knowledge graph in the specific field according to the specific legal decision book text.
Example (b):
first, scene description
Illegal card swiping court judgment:
the court trial found out: wang XX transacts the procedure of opening an account in Z bank in 2015, 4 months and 22 days, and applies for a savings card. This card finds XXX by ATM in nation T at 13 pm 32 minutes 18 seconds on 1 st pm of 9/2015. The method comprises the following steps that the king XX receives a bank expense short message at home at 35 minutes and 10 seconds at 13 pm on 1 st/9/2015, gives an alarm immediately and goes to a nearby counter to handle a loss report procedure, and the specific time is as follows: the phone call of the Z bank is lost when the phone call is made for 36 minutes and 01 seconds at 13 o 'clock and 57 o' clock of the Z bank in 2015, 9, 1, 13 o 'clock and 57 o' clock of the Z bank, and then the lost call procedure is handled at 23 o 'clock of the Z bank at 14 o' clock. Additionally, it can be found that there is no record of the eminence in 2015 year.
Second, the concrete steps
Step 1: referring to fig. 2, a domain expert constructs a legal knowledge map of the illegal card swiping according to the judgment text of the illegal card swiping court;
step 2: setting a predicate-oriented lexicon library by extracting predicate verbs of triples in the knowledge graph stolen by the pseudo-card, such as account opening, card swiping, alarming, loss reporting and the like;
and step 3: selecting different knowledge maps and predicate-oriented word banks according to the types of the new judgment books;
and 4, step 4: setting the starting position and the ending position of a judgment book text, wherein if the starting position is found out in court trial, the ending position is considered in the court;
and 5: referring to FIG. 3, the sentences in the start and stop positions are numbered in sequence a1、…、anAnd performing part-of-speech tagging;
step 6: matching sentences in the new judgment book text one by using the predicate guide words;
and 7: establishing a triple of the sentences matched with the predicate guide words, and adding the triple into the existing knowledge graph;
and 8: for sentences "b" not matched to predicate directors1: additionally, find out the label of "there is no exit record in 2015 year;
and step 9: to b is1Screening, summarizing a new predicate verb 'exit', and judging whether to add the new predicate verb 'exit' into a predicate-oriented lexicon; if a new predicate director is put in storage, the predicate director is marked as C1: outbound ", and selecting a judgment book text for constructing a knowledge graph;
step 10: performing the operations of step 4 to step 6 on the existing judgment books by using the new predicate guide;
step 11: repeating the step 3 to the step 10 when a new judgment book text appears;
step 12: referring to fig. 4, the completion of the domain-specific knowledge graph is performed according to a specific decision text.
The above embodiments are only used for illustrating the design idea and features of the present invention, and the purpose of the present invention is to enable those skilled in the art to understand the content of the present invention and implement the present invention accordingly, and the protection scope of the present invention is not limited to the above embodiments. Therefore, all equivalent changes and modifications made in accordance with the principles and concepts disclosed herein are intended to be included within the scope of the present invention.
Claims (10)
1. A knowledge graph completion method for automatically extracting relations based on a decision book text is characterized by comprising the following steps: the method comprises the following steps:
s01: constructing a legal knowledge graph according to the types of the judgment book texts;
s02: extracting the relation in the legal knowledge graph;
s06: extracting the relation of sentences in the text of the decision book according to the relation obtained in the step S02, and executing a step S07 if the extraction is successful; if the extraction is not possible, go to step S08;
s07: adding the relationship extracted in step S06 to the legal knowledge base;
s08: the relationship that could not be extracted in step S06 is added to the predicate-oriented thesaurus and the legal knowledge graph is complemented.
2. The knowledge-graph completion method for automatic extraction relationship based on decision-making text as claimed in claim 1, wherein: in the step S02, the specific steps are as follows: and extracting predicate verbs of the triples in the legal knowledge graph and setting a predicate-oriented lexicon.
3. The knowledge-graph completion method for automatic extraction relationship based on decision-making text as claimed in claim 2, wherein: in the step S06, the specific steps are as follows: the predicate-directors in the predicate-director lexicon are used to match the sentences in the decision-book text one by one.
4. The knowledge-graph completion method for automatic extraction relationship based on decision-making text as claimed in claim 3, wherein: in the step S07, the specific steps are as follows: and establishing a triple through the sentences matched with the predicate guidance words, and adding the triple into the legal knowledge graph.
5. The knowledge-graph completion method for automatic extraction relationship based on decision-making text as claimed in claim 4, wherein: in the step S08, the specific steps are as follows: and screening sentences which are not matched with the predicate guide words, adding new predicate verbs into a predicate guide word library, and completing the legal knowledge graph.
6. The knowledge-graph completion method for automatic extraction relationship based on decision-making text as claimed in claim 1, wherein: between the step S02 and the step S06, the method further includes the following steps:
s03: selecting different legal knowledge maps and predicate guide lexicons according to the types of the judgment book texts;
s04: setting a starting position and an ending position of a judgment book text;
s05: the sentences between the starting position and the ending position in the text of the judgment book are sequentially numbered as a1、…、anAnd parts of speech are labeled to the components of the sentence.
7. The knowledge-graph completion method for automatic extraction relationship based on decision-making text as claimed in claim 6, wherein: in step S08, sentences that do not match the predicate director are sequentially numbered as b1、…、bn(ii) a Sequentially numbering new predicate directors added into the predicate-oriented lexicon as C1、…、Cn。
8. The knowledge-graph completion method for automatic extraction relationship based on decision-making text as claimed in claim 1, wherein: further comprising the steps of:
s09: and selecting the judgment text for constructing the legal knowledge graph, and executing the steps S04 to S06 on the existing judgment text by using the new predicate guide words.
9. The knowledge-graph completion method for automatic extraction relationship based on decision-making text as claimed in claim 1, wherein: further comprising the steps of:
s10: executing the step S03 to the step S09 to the newly added judgment book text;
s11: and completing the legal knowledge map according to the newly added judgment book text.
10. A computer storage medium, characterized in that: stored therein is a computer program executable by a computer processor, the computer program performing a method of knowledge-graph completion based on decision-text automatic extraction relations as claimed in any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110478354.7A CN113204648A (en) | 2021-04-30 | 2021-04-30 | Knowledge graph completion method based on automatic extraction relationship of judgment book text |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110478354.7A CN113204648A (en) | 2021-04-30 | 2021-04-30 | Knowledge graph completion method based on automatic extraction relationship of judgment book text |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113204648A true CN113204648A (en) | 2021-08-03 |
Family
ID=77028733
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110478354.7A Pending CN113204648A (en) | 2021-04-30 | 2021-04-30 | Knowledge graph completion method based on automatic extraction relationship of judgment book text |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113204648A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117743601A (en) * | 2024-02-05 | 2024-03-22 | 中南大学 | Natural resource knowledge graph completion method, device, equipment and medium |
| CN117743601B (en) * | 2024-02-05 | 2024-05-17 | 中南大学 | Natural resource knowledge graph completion method, device, equipment and medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180144252A1 (en) * | 2016-11-23 | 2018-05-24 | Fujitsu Limited | Method and apparatus for completing a knowledge graph |
| CN110309321A (en) * | 2019-07-10 | 2019-10-08 | 电子科技大学 | A kind of representation of knowledge learning method indicating study based on map |
| CN110427496A (en) * | 2019-08-05 | 2019-11-08 | 北京百度网讯科技有限公司 | Knowledge mapping extending method and device for text-processing |
| CN111159385A (en) * | 2019-12-31 | 2020-05-15 | 南京烽火星空通信发展有限公司 | Template-free universal intelligent question-answering method based on dynamic knowledge graph |
| CN111209412A (en) * | 2020-02-10 | 2020-05-29 | 同方知网(北京)技术有限公司 | Method for building knowledge graph of periodical literature by cyclic updating iteration |
| CN112612899A (en) * | 2020-11-24 | 2021-04-06 | 中国传媒大学 | Knowledge graph construction method and device, storage medium and electronic equipment |
-
2021
- 2021-04-30 CN CN202110478354.7A patent/CN113204648A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180144252A1 (en) * | 2016-11-23 | 2018-05-24 | Fujitsu Limited | Method and apparatus for completing a knowledge graph |
| CN110309321A (en) * | 2019-07-10 | 2019-10-08 | 电子科技大学 | A kind of representation of knowledge learning method indicating study based on map |
| CN110427496A (en) * | 2019-08-05 | 2019-11-08 | 北京百度网讯科技有限公司 | Knowledge mapping extending method and device for text-processing |
| CN111159385A (en) * | 2019-12-31 | 2020-05-15 | 南京烽火星空通信发展有限公司 | Template-free universal intelligent question-answering method based on dynamic knowledge graph |
| CN111209412A (en) * | 2020-02-10 | 2020-05-29 | 同方知网(北京)技术有限公司 | Method for building knowledge graph of periodical literature by cyclic updating iteration |
| CN112612899A (en) * | 2020-11-24 | 2021-04-06 | 中国传媒大学 | Knowledge graph construction method and device, storage medium and electronic equipment |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117743601A (en) * | 2024-02-05 | 2024-03-22 | 中南大学 | Natural resource knowledge graph completion method, device, equipment and medium |
| CN117743601B (en) * | 2024-02-05 | 2024-05-17 | 中南大学 | Natural resource knowledge graph completion method, device, equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Qian et al. | Towards automated reentrancy detection for smart contracts based on sequential models | |
| CN110321432B (en) | Text event information extraction method, electronic device and nonvolatile storage medium | |
| US10565244B2 (en) | System and method for text categorization and sentiment analysis | |
| Forsythand et al. | Lexical and discourse analysis of online chat dialog | |
| CN105868179B (en) | A kind of intelligent answer method and device | |
| US20170228365A1 (en) | Systems and methods for language feature generation over multi-layered word representation | |
| US11847423B2 (en) | Dynamic intent classification based on environment variables | |
| CN110555206A (en) | named entity identification method, device, equipment and storage medium | |
| CN108710704A (en) | Determination method, apparatus, electronic equipment and the storage medium of dialogue state | |
| US20210073336A1 (en) | Progressive collocation for real-time discourse | |
| CN111859968A (en) | Text structuring method, text structuring device and terminal equipment | |
| CN111738018A (en) | Intention understanding method, device, equipment and storage medium | |
| CN112132238A (en) | Method, device, equipment and readable medium for identifying private data | |
| CN111143571A (en) | Entity labeling model training method, entity labeling method and device | |
| CN113343677B (en) | Intention identification method and device, electronic equipment and storage medium | |
| US11934787B2 (en) | Intent determination in a messaging dialog manager system | |
| US20220272124A1 (en) | Using machine learning for detecting solicitation of personally identifiable information (pii) | |
| CN109672586A (en) | A kind of DPI service traffics recognition methods, device and computer readable storage medium | |
| CN111563140B (en) | Intention identification method and device | |
| CN113204648A (en) | Knowledge graph completion method based on automatic extraction relationship of judgment book text | |
| CN109684357A (en) | Information processing method and device, storage medium, terminal | |
| CN115374259A (en) | Question and answer data mining method and device and electronic equipment | |
| Yadav et al. | Sentiment analysis of financial news using unsupervised and supervised approach | |
| CN113792140A (en) | Text processing method and device and computer readable storage medium | |
| CN111401034A (en) | Text semantic analysis method, semantic analysis device and terminal |
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 |
Application publication date: 20210803 |
|
| RJ01 | Rejection of invention patent application after publication |