CN115964471A - Approximate query method for medical data - Google Patents
Approximate query method for medical data Download PDFInfo
- Publication number
- CN115964471A CN115964471A CN202310255574.2A CN202310255574A CN115964471A CN 115964471 A CN115964471 A CN 115964471A CN 202310255574 A CN202310255574 A CN 202310255574A CN 115964471 A CN115964471 A CN 115964471A
- Authority
- CN
- China
- Prior art keywords
- query
- approximate
- model
- gpt
- question
- 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 26
- 239000011159 matrix material Substances 0.000 claims description 10
- 230000006870 function Effects 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 8
- 238000007476 Maximum Likelihood Methods 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 230000002708 enhancing effect Effects 0.000 abstract description 2
- 230000002776 aggregation Effects 0.000 description 2
- 238000004220 aggregation Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention belongs to the technical field of database query, and discloses a medical data approximate query method, which comprises the following steps: converting the approximate query records into natural language form representation, and performing data enhancement on the query problems in the approximate query records in the natural language form by using a synonym replacement mode; enriching the query result in the approximate query record in the natural language form; combining the query question and the query result into a plurality of question-answer pairs, and forming a question-answer set by the plurality of question-answer pairs; fine-tuning the GPT-3 model by using a question-answer set; and inputting the natural language query into the GPT-3 model and outputting an answer result. The medical data approximate query method uses a GPT-3 model to realize medical data query with ultralow access delay, and enriches query results by enhancing data of query problems; the GPT-3 model is better suitable for approximate access of the database, and the accuracy of the approximate access is improved.
Description
Technical Field
The invention belongs to the technical field of database query, and particularly relates to a medical data approximate query method based on a GPT-3 model.
Background
The approximate query processing is a key problem in a database, the approximate query processing refers to an optimization technology for quickly responding to the query efficiency of a user under an acceptable query error, compared with the traditional database query, the approximate query can greatly improve the query speed of the database under the condition of slightly sacrificing the query precision, and the approximate query processing is usually applied to a commercial database with a large data volume and mainly aims at query statements containing aggregation operations such as 'count, sum, avg' and the like.
In the field of approximate queries, the prior art focuses on improving query optimizers in databases to ensure that more efficient execution plans can be compiled for approximate query statements, thereby speeding up the overall data query process. However, in the process of implementing the present invention, the inventor finds that at least the following problems exist in the prior art:
1. query efficiency is still difficult to achieve in milliseconds. Although the existing approximate query technology can reduce the data range to be queried by sampling and other methods, in the case of a large amount of data, the process is still extremely time-consuming. Secondly, the prior art focuses on improving a query optimizer inside a database to accelerate the query, but still needs to execute a standard query statement in the database, which cannot avoid time-consuming operations such as data scanning and transmission;
2. an approximate query in natural language cannot be answered. Usually, a database provider needs a user to have a certain SQL writing capability, and even some users master the writing of simple SQL in time, the execution speed is still too slow because the written SQL is not standard enough. This results in a high requirement of the database provider for the expertise of the user.
Disclosure of Invention
The present invention aims to solve the above technical problem at least to some extent. Therefore, the invention aims to provide a medical data approximate query method.
The technical scheme adopted by the invention is as follows:
the approximate query method of the medical data comprises the following steps:
s1, converting the approximate query records into a representation in a natural language form through a Transformer model, and performing data enhancement on query problems in the approximate query records in the natural language form by using a synonym replacement mode; enriching the query result in the approximate query record in the natural language form; combining the query questions subjected to data enhancement and corresponding query results into a plurality of question-answer pairs, and forming a question-answer set by the plurality of question-answer pairs;
s2, processing the question and answer set into a data format comprising prompts and conclusions, and calling a fine-tuning API of the GPT-3 model by using the processed question and answer set to fine-tune the GPT-3 model;
and S3, inputting the natural language query into the fine-tuned GPT-3 model, and outputting an answer result of the GPT-3 model.
Preferably, the approximate query record in step S1 includes a historical approximate query record and a randomly generated approximate query record, and the randomly generated approximate query record is randomly generated by fixing the query template.
Preferably, the Transformer model in step S1 represents the query language as a two-dimensional matrix X 1 Wherein each vector represents the embedded representation of each participle (token) in the query language and is converted into three matrices of Q, K and V by three linear changes, and the conversion formula is as follows:
wherein, matMul represents linear matrix multiplication,
representing three different two-dimensional matrices, respectively.
Preferably, the objective function of the GPT-3 model fine tuning process is constructed by a maximum likelihood function:
wherein ,
represents a question in a question-answer pair, and->
The first i words that represent the query result, and θ represents the parameters of the GPT-3 model.
Preferably, the GPT-3 model in step S3 accepts a matrix representing a natural language query
And outputs a word y 0 And then->
The second word is output for input and so on until the end identifier is output.
The invention has the beneficial effects that:
the medical data approximate query method provided by the invention realizes medical data query with ultralow access delay by using a GPT-3 model, and enriches query results in a natural language form by enhancing data of query problems; therefore, the GPT-3 model is better suitable for approximate access of the database, and the accuracy of the approximate access is improved.
The medical data approximate query method also solves the problem of insufficient historical approximate query records through randomly generating the approximate query records in a fixed query template mode.
Drawings
FIG. 1 is a flow chart of a method of approximate query of medical data of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it should be understood that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should also be noted that, in some embodiments, the functions/acts noted may occur out of the order noted in the figures. For example, two figures shown in succession may, in fact, be executed substantially concurrently, or the figures may sometimes be executed in the reverse order, depending upon the functionality/acts involved.
As shown in fig. 1, the medical data approximate query method of the embodiment includes the following steps:
s1, converting the approximate query records into a representation in a natural language form through a Transformer model, wherein the Transformer model has the advantage of performing parallel computation by representing input as a matrix compared with a sequence model, so that the training and reasoning efficiency is improved. The Transformer model represents the inputs as a two-dimensional matrix X 1 And the three linear changes are converted into three matrixes of Q, K and V, so that the self-attention mechanism can be facilitated. The conversion formula is:
wherein, matMul represents linear matrix multiplication,
representing three different two-dimensional matrices, respectively.
The approximate query records comprise historical approximate query records and randomly generated approximate query records, and insufficient historical approximate query records exist in the target medical database, namely the historical approximate query records are insufficient to cover various data ranges (attributes and attribute values) and query types ("count, sum, avg"), and the historical approximate query records need to be supplemented by the randomly generated approximate query records generated randomly by means of a fixed query template. For example, if the data range of the existing large amount of historical approximate query records is "number of patients", then the randomly generated approximate query records with the data range of "number of patient hospitalizations" or other data ranges need to be randomly generated, and by combining the randomly generated approximate query records and the historical approximate query records, sufficient fine-tuning data can be formed for the subsequent steps.
And performing data enhancement on the query problem in the approximate query record in the natural language form by using a synonym replacement mode to strengthen the generalization capability of the GPT-3 model after fine adjustment. The process retrieves and replaces words or terms in the query question according to an open source synonym table (such as a hayada synonym table) and stores as a new query. For example, for the query question "how many patients older than 50 are", the synonym "total" is queried in the synonym table using "number", and the new query after replacement is "how many patients older than 50 are".
Enriching the query result in the approximate query record in the natural language form, and finely adjusting the GPT-3 model together with the query problem in the natural language form. For example, if the query question is "how many patients older than 50 years" and the query result is "2003", the query result is enriched to "2003". The process automatically converts the query results to natural language based on the type of aggregation operation (SUM, count..) of the query, and the name of the attribute being queried (such as "patient"). Specifically, it is necessary to capture "SUM (patient)" in the query by regular expression, translate "SUM" into "total number", and extract the attribute name of "patient". The "+" aggregate operation (total) "+" finally associated with query result "2003" in accordance with "attribute (patient)" + "is" + "query result (2003)" + ". "sequential combination of natural language" the total number of patients was 2003. "is used.
And combining each query question subjected to data enhancement and the corresponding query result into a plurality of question-answer pairs, and forming a question-answer set by the plurality of question-answer pairs.
S2, processing the question and answer set into a data format comprising prompts and conclusions, wherein the data format is { "prompt": < prompt text > "," completion ": ideal generated text > }; and calling a fine-tuning API of the GPT-3 model by using the processed question-answer set to fine-tune the GPT-3 model, so that the GPT-3 model can accurately answer the approximate query aiming at the target medical database. The GPT-3 model is adopted because the pre-trained GPT-3 model has better learning effect under a small sample.
The formula of the GPT-3 model is as follows:
wherein, Q represents input information, which is information existing in the input text. K represents content information, namely semantic information, attentition (Q, K) represents the matching degree of Query and Key, and V represents information per se and has the main function of weighting the matching degree;
for a calculation result of multi-head attention, concat represents matrix splicing; the target function of the GPT-3 model fine tuning process is constructed by a maximum likelihood function:
wherein ,
represents a question in a question-answer pair, and->
The first i words representing the query results, and θ represents the parameters of the GPT-3 model.
And S3, inputting the natural language query of the user into the finely-adjusted GPT-3 model, and returning the answer result of the GPT-3 model to the user. In particular, the GPT-3 modelAccepting natural language queries
And outputs a word y 0 And then->
The second word is output for input and so on until the end identifier is output. Such as a user giving a query in natural language: the number of people who catch the cold each month. ", the GPT-3 model would accept the statement and output the first word" number "most likely to appear in the answer, after which the GPT-3 model accepts" the number of people that are cold monthly. Number ", and output" amount "; and by analogy, the number of the complete outputs is finally 500.".
The medical data approximate query method can obtain corresponding answers only by describing an approximate query by natural language without a user mastering an efficient SQL compiling technology, thereby reducing the requirement of a medical database provider on professional knowledge of the user; meanwhile, medical data query with ultra-low access delay can be realized. When a user inputs a natural language query, the natural language query is matched with the historical records and a query result similar to the natural language query is quickly returned without actually executing the standard SQL query in the database. This greatly improves the reply efficiency of the query (up to the millisecond ms level).
The invention is not limited to the above alternative embodiments, and any other various forms of products can be obtained by anyone in the light of the present invention, but any changes in shape or structure thereof, which fall within the scope of the present invention as defined in the claims, fall within the scope of the present invention.
Claims (5)
1. A medical data approximate query method is characterized by comprising the following steps:
s1, converting the approximate query records into a representation in a natural language form through a Transformer model, and performing data enhancement on query problems in the approximate query records in the natural language form by using a synonym replacement mode; enriching the query result in the approximate query record in the natural language form; combining the query questions subjected to data enhancement and corresponding query results into a plurality of question-answer pairs, and forming a question-answer set by the plurality of question-answer pairs;
s2, processing the question and answer set into a data format comprising prompts and conclusions, and calling a fine-tuning API of the GPT-3 model by using the processed question and answer set to fine-tune the GPT-3 model;
and S3, inputting the natural language query into the fine-tuned GPT-3 model, and outputting an answer result of the GPT-3 model.
2. The approximate query method for medical data as set forth in claim 1, wherein: the approximate query records in the step S1 comprise historical approximate query records and randomly generated approximate query records, and the randomly generated approximate query records are randomly generated in a mode of fixing a query template.
3. The approximate query method for medical data as set forth in claim 1, wherein: in step S1, the Transformer model expresses the query language as a two-dimensional matrix X 1 Wherein each vector represents the embedded representation of each participle in the query language and is converted into three matrices of Q, K and V by three linear changes, the conversion formula is:
wherein, matMul represents linear matrix multiplication,
representing three different two-dimensional matrices, respectively.
4. The approximate query method for medical data as claimed in claim 1, wherein: the target function of the GPT-3 model fine tuning process is constructed by a maximum likelihood function:
wherein ,
represents a question in a question-answer pair, and->
The first i words that represent the query result, and θ represents the parameters of the GPT-3 model.
5. The approximate query method for medical data as set forth in claim 1, wherein: step S3 the GPT-3 model accepts a matrix representing a natural language query
And outputs a word y 0 And then->
The second word is output for input, and so on until the end identifier is output. />
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310255574.2A CN115964471B (en) | 2023-03-16 | 2023-03-16 | Medical data approximate query method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310255574.2A CN115964471B (en) | 2023-03-16 | 2023-03-16 | Medical data approximate query method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115964471A true CN115964471A (en) | 2023-04-14 |
| CN115964471B CN115964471B (en) | 2023-06-02 |
Family
ID=85888171
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310255574.2A Active CN115964471B (en) | 2023-03-16 | 2023-03-16 | Medical data approximate query method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115964471B (en) |
Citations (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080010259A1 (en) * | 2006-07-10 | 2008-01-10 | Nec (China) Co., Ltd. | Natural language based location query system, keyword based location query system and a natural language and keyword based location query system |
| US20100070521A1 (en) * | 2008-09-18 | 2010-03-18 | Xerox Corporation | Query translation through dictionary adaptation |
| US20120254143A1 (en) * | 2011-03-31 | 2012-10-04 | Infosys Technologies Ltd. | Natural language querying with cascaded conditional random fields |
| CN103218436A (en) * | 2013-04-17 | 2013-07-24 | 中国科学院自动化研究所 | Similar problem retrieving method fusing user category labels and device thereof |
| US20160239487A1 (en) * | 2015-02-12 | 2016-08-18 | Microsoft Technology Licensing, Llc | Finding documents describing solutions to computing issues |
| WO2017010652A1 (en) * | 2015-07-15 | 2017-01-19 | 포항공과대학교 산학협력단 | Automatic question and answer method and device therefor |
| US20170228372A1 (en) * | 2016-02-08 | 2017-08-10 | Taiger Spain Sl | System and method for querying questions and answers |
| CN109766355A (en) * | 2018-12-28 | 2019-05-17 | 上海汇付数据服务有限公司 | A kind of data query method and system for supporting natural language |
| WO2020034642A1 (en) * | 2018-08-17 | 2020-02-20 | 齐鲁工业大学 | Automatic medical question answering method and apparatus, storage medium, and electronic device |
| CN111737399A (en) * | 2020-05-28 | 2020-10-02 | 北京百度网讯科技有限公司 | Method and device for expanding question and answer set, electronic equipment and readable storage medium |
| CN111813802A (en) * | 2020-09-11 | 2020-10-23 | 杭州量之智能科技有限公司 | Method for generating structured query statement based on natural language |
| CN112035620A (en) * | 2020-08-31 | 2020-12-04 | 康键信息技术（深圳）有限公司 | Question-answer management method, device, equipment and storage medium of medical query system |
| CN112905868A (en) * | 2021-03-22 | 2021-06-04 | 京东方科技集团股份有限公司 | Event extraction method, device, equipment and storage medium |
| CN113297364A (en) * | 2021-06-07 | 2021-08-24 | 吉林大学 | Natural language understanding method and device for dialog system |
| US20210294970A1 (en) * | 2020-03-23 | 2021-09-23 | Sorcero, Inc. | Cross-class ontology integration for language modeling |
| CN113553412A (en) * | 2021-06-30 | 2021-10-26 | 北京百度网讯科技有限公司 | Question and answer processing method and device, electronic equipment and storage medium |
| CN114897163A (en) * | 2022-05-23 | 2022-08-12 | 阿里巴巴（中国）有限公司 | Pre-training model data processing method, electronic device and computer storage medium |
| US20220414737A1 (en) * | 2021-06-28 | 2022-12-29 | Microsoft Technology Licensing, Llc | Query-based product representations |
| CN115587583A (en) * | 2022-11-07 | 2023-01-10 | 维沃移动通信有限公司 | Noise detection method and device and electronic equipment |
-
2023
- 2023-03-16 CN CN202310255574.2A patent/CN115964471B/en active Active
Patent Citations (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080010259A1 (en) * | 2006-07-10 | 2008-01-10 | Nec (China) Co., Ltd. | Natural language based location query system, keyword based location query system and a natural language and keyword based location query system |
| US20100070521A1 (en) * | 2008-09-18 | 2010-03-18 | Xerox Corporation | Query translation through dictionary adaptation |
| US20120254143A1 (en) * | 2011-03-31 | 2012-10-04 | Infosys Technologies Ltd. | Natural language querying with cascaded conditional random fields |
| CN103218436A (en) * | 2013-04-17 | 2013-07-24 | 中国科学院自动化研究所 | Similar problem retrieving method fusing user category labels and device thereof |
| US20160239487A1 (en) * | 2015-02-12 | 2016-08-18 | Microsoft Technology Licensing, Llc | Finding documents describing solutions to computing issues |
| WO2017010652A1 (en) * | 2015-07-15 | 2017-01-19 | 포항공과대학교 산학협력단 | Automatic question and answer method and device therefor |
| US20170228372A1 (en) * | 2016-02-08 | 2017-08-10 | Taiger Spain Sl | System and method for querying questions and answers |
| WO2020034642A1 (en) * | 2018-08-17 | 2020-02-20 | 齐鲁工业大学 | Automatic medical question answering method and apparatus, storage medium, and electronic device |
| CN109766355A (en) * | 2018-12-28 | 2019-05-17 | 上海汇付数据服务有限公司 | A kind of data query method and system for supporting natural language |
| US20210294970A1 (en) * | 2020-03-23 | 2021-09-23 | Sorcero, Inc. | Cross-class ontology integration for language modeling |
| CN111737399A (en) * | 2020-05-28 | 2020-10-02 | 北京百度网讯科技有限公司 | Method and device for expanding question and answer set, electronic equipment and readable storage medium |
| CN112035620A (en) * | 2020-08-31 | 2020-12-04 | 康键信息技术（深圳）有限公司 | Question-answer management method, device, equipment and storage medium of medical query system |
| CN111813802A (en) * | 2020-09-11 | 2020-10-23 | 杭州量之智能科技有限公司 | Method for generating structured query statement based on natural language |
| CN112905868A (en) * | 2021-03-22 | 2021-06-04 | 京东方科技集团股份有限公司 | Event extraction method, device, equipment and storage medium |
| CN113297364A (en) * | 2021-06-07 | 2021-08-24 | 吉林大学 | Natural language understanding method and device for dialog system |
| US20220414737A1 (en) * | 2021-06-28 | 2022-12-29 | Microsoft Technology Licensing, Llc | Query-based product representations |
| CN113553412A (en) * | 2021-06-30 | 2021-10-26 | 北京百度网讯科技有限公司 | Question and answer processing method and device, electronic equipment and storage medium |
| CN114897163A (en) * | 2022-05-23 | 2022-08-12 | 阿里巴巴（中国）有限公司 | Pre-training model data processing method, electronic device and computer storage medium |
| CN115587583A (en) * | 2022-11-07 | 2023-01-10 | 维沃移动通信有限公司 | Noise detection method and device and electronic equipment |
Non-Patent Citations (7)
| Title |
|---|
| ANIU: "transformers：最先进的自然语言处理", pages 1 - 2, Retrieved from the Internet <URL:https://zhuanlan.zhihu.com/p/374868051> * |
| JACEK KANIEWSKI 等: "Hybrid Transformer With Matrix Converter", IEEE TRANSACTIONS ON POWER DELIVERY, pages 1388 * |
| SEN HU 等: "Answering Natural Language Questions by Subgraph Matching over Knowledge Graphs", IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING, pages 824 * |
| 刘紫琪: "面向科技人才信息问答的查询需求及答案组织研究", 文献与数据学报, pages 30 - 43 * |
| 张帅: "海量数据近似查询处理技术研究", 中国优秀硕士学位论文全文数据库 信息科技辑, no. 4, pages 138 - 354 * |
| 铁四角: "近似查询mysql_近似查询处理简介", pages 1, Retrieved from the Internet <URL:https://blog.csdn.net/weixin_34461414/article/details/113713297> * |
| 顾迎捷 等: "基于神经网络的机器阅读理解综述", 软件学报, no. 07, pages 2095 - 2120 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115964471B (en) | 2023-06-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2020010834A1 (en) | Faq question and answer library generalization method, apparatus, and device | |
| CN116701431A (en) | Data retrieval method and system based on large language model | |
| CN111400453B (en) | Intelligent interaction system and method based on legal consultation | |
| CN117370580A (en) | Knowledge-graph-based large language model enhanced dual-carbon field service method | |
| CN111625554A (en) | Data query method and device based on deep learning semantic understanding | |
| CN116932776A (en) | Knowledge graph-based large model knowledge updating method and device | |
| Wang et al. | Exploring New Frontiers of Deep Learning in Legal Practice: A Case Study of Large Language Models | |
| CN111597316A (en) | Multi-stage attention answer selection method fusing semantics and question key information | |
| CN112632106B (en) | Knowledge graph query method, device, equipment and storage medium | |
| CN115964471B (en) | Medical data approximate query method | |
| Trummer | Demonstrating the voice-based exploration of large data sets with CiceroDB-zero | |
| CN116932717A (en) | Knowledge base retrieval method and device based on large language model self-evaluation and self-feedback | |
| CN115964468A (en) | Rural information intelligent question-answering method and device based on multilevel template matching | |
| CN114817510A (en) | Question and answer method, question and answer data set generation method and device | |
| Wang et al. | A multiple-integration encoder for multi-turn text-to-SQL semantic parsing | |
| CN112015920A (en) | Intelligent auxiliary learning system based on knowledge graph and edge calculation | |
| CN118170894A (en) | Knowledge graph question-answering method, knowledge graph question-answering device and storage medium | |
| CN117668179A (en) | Financial index accurate question-answering method based on large model | |
| Holsapple | Framework for a generalized intelligent decision support system. | |
| Mincheva et al. | NLP using database context | |
| CN118152427A (en) | SQL analysis method, system, terminal and medium based on large model | |
| CN118051594A (en) | SQL sentence processing model construction and SQL sentence generation method and device | |
| Yang | Research on the Intelligent Question Answering based on knowledge graph | |
| CN117909458A (en) | Construction method of mould specialized question-answering system based on LLM model | |
| CN117113971A (en) | Data set generation method and system based on natural language analysis |
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 | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Liu Ruihua Inventor after: Zhang Jintao Inventor after: Li Rui Inventor after: Hu Qitong Inventor after: Zheng Mingyang Inventor before: Liu Ruihua Inventor before: Zhang Jintao Inventor before: Li Rui Inventor before: Hu Qitong Inventor before: Zheng Mingyang Inventor before: Tang Xuewen |