CN107368547A - A kind of intelligent medical automatic question-answering method based on deep learning - Google Patents
A kind of intelligent medical automatic question-answering method based on deep learning Download PDFInfo
- Publication number
- CN107368547A CN107368547A CN201710510408.7A CN201710510408A CN107368547A CN 107368547 A CN107368547 A CN 107368547A CN 201710510408 A CN201710510408 A CN 201710510408A CN 107368547 A CN107368547 A CN 107368547A
- Authority
- CN
- China
- Prior art keywords
- answer
- question
- vector
- text
- medical
- 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
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
- G06F16/374—Thesaurus
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The invention discloses a kind of intelligent medical automatic question-answering method based on deep learning, including:1) answer pond is built, then vectorization is carried out to the data in the answer pond;2) medical question and answer data are obtained from internet, then vectorization is carried out to medical question and answer data, obtain training text vector, then its vector is input in neutral net, and model training cost function is built, then the neutral net is trained, the neutral net after must training;3) term vector of each word in waiting to answer a question is extracted, then the term vector of each word is spliced, corresponding text vector of answering a question must be waited;4) step 3) is obtained in the neutral net after text vector is input to training, obtain the characteristic vector of text, the characteristic vector of text is input in answer pond again, obtain the characteristic vector of some candidate answers, then optimum answer is chosen, and the optimum answer is exported, the answering method can realize automatic, the accurate answer of medical care problem.
Description
Technical field
The invention belongs to domain-oriented automatically request-answering system technical field, is related to a kind of intelligent medical based on deep learning
Automatic question-answering method.
Background technology
With the development of science and technology with the progress of society, the community's question answering system for being related to medical field have accumulated sea at present
Measure data.But retrieval is difficult, information is very different to be difficult to meet that the popular online question and answer consulting of " internet+" epoch is particularly medical
The demand of question and answer consulting.
Existing online healthcare community question and answer website is such as sought medical advice and medicine net at present, though 99 healthy nets are big after the accumulation of more than ten years
Amount data but user can only be manually browsed using retrieval mode from nearly ten million data needed for lookup, and if user proposes problem
Then medical practitioner is waited to answer, and the stand-by period is generally at several days or so.And doctor generally requires to ask same or analogous
Topic is answered repeatedly, is greatly wasted manpower and materials, has been delayed patient's therapic opportunity.And other answer efficiency higher medical treatment
Community's answer platform such as spring rain doctor, then need to collect more enquirement expense, by APP line medical practitioners, seek advice from sometimes
Expense has been even more than goes to hospital register the expenses of diagnosis and treatment using public health insurance, loses convenience and economy.
A variety of factors cause great difficulty to online medical question and answer, consulting.
Automatic question answering is an important direction of natural language processing field, it is intended to allows user directly to be carried with natural language
Ask and obtain answer.For example, " where is Baidu mansion for user's query", question answering system answers the " street of ShangDi, Haidian District, BeiJing City ten
No. 10 ".From the point of view of user, automatic question answering is a kind of simple and succinct information acquisition method.User directly uses nature language
Speech interacts with question answering system, and without considering which type of keyword combination to represent the intention of oneself using, so simply;Question and answer
System directly returns to the answer of problem, and user from tediously long relevant documentation without finding answer content, so succinctly.Traditional
Automatic question answering technology mainly includes three parts:Case study, information retrieval and answer extracting.The purpose of case study is analysis
The semantic type of problem, the intention that user puts question to is determined, i.e. user is query time, place or inquiry entity, entity attribute
Or other, and the keyword in extraction problem;Information retrieval is the keyword and its extension form obtained according to case study
Relevant documentation is retrieved from online or offline document library;Answer extracting is to extract to answer in the relevant documentation obtained from retrieval
Case.According to the difference of problem types, answer form is also not quite similar, it may be possible to a word, a sentence, it is also possible to one
Paragraph or longer text string.
Automatic question answering technology is divided into three types, is respectively:Traditional question and answer technology, the automatic question answering of knowledge based collection of illustrative plates
Technology is with the identification of keyword and parsing and problem with question and answer technology traditional the automatic question answering technology based on deep learning
Method of completing the square is core, the sequence to candidate answers, but this mode efficiency is very low, the very poor knowledge baseds collection of illustrative plates of autgmentability
Automatic question answering technology, be the concept that knowledge mapping is added in the method based on traditional question and answer technology, pass through knowledge rope
The mode drawn carries out automatic question answering in this approach, and the link of most critical is to utilize Entity recognition and entity-relationship recognition skill
This technologies of structure that art carries out knowledge mapping are clearly disadvantageous there is some:First, knowledge mapping is by objectively real
Body relation structure, more be answer some facts thing, the problem of some subjectivities, is answered it is not accurate enough, therefore
Often effect is bad in community's question and answer;Secondly, when needing to solve towards the automatic question answering problem of wide-range and mass data,
The structure of knowledge mapping is sufficiently complex, and if also be difficult to be guaranteed the knowledge of specific areas for the accuracy of structure
Gaps and omissions be present, then can not obtain the answer of association area.
At present, China's medical level improves constantly, and common people's health care consciousness constantly enhancing, most of patient know in time
The importance of medical treatment.But increasing for medical treatment number makes originally limited medical resource increasingly have too many difficulties to cope with.No matter state of an illness weight
Emergency, the subconscious large hospital for pouring into provincial capital of many patients, causes the waste of primary care resource, quick-fried with large hospital
Full, many critical patients cannot be treated timely on the contrary.At the same time, doctor-patient relationship is continuous worsening, the problems such as medical trouble also gradually
Gradually become serious social concern.
The content of the invention
A kind of the shortcomings that it is an object of the invention to overcome above-mentioned prior art, there is provided intelligence doctor based on deep learning
Automatic question-answering method is treated, the answering method can realize automatic, the accurate answer of medical care problem.
To reach above-mentioned purpose, the intelligent medical automatic question-answering method of the present invention based on deep learning includes following
Step:
1) structure is stored with the answer pond of medical care problem, the answer of the medical care problem and the medical care problem art,
Vectorization is carried out to the data in the answer pond again;
2) medical question and answer data are obtained from internet, then the medical question and answer data to acquiring carry out vectorization, obtain
The training text vector, is then input in neutral net by training text vector, and builds model training cost function, then
The neutral net is trained by model training cost function, the neutral net after must training;
3) treat to answer a question and carry out vectorization, then extract wait to answer a question in each word term vector, then will treat
The term vector of each word is spliced in answering a question, and must wait corresponding text vector of answering a question;
4) in the neutral net after text vector corresponding to answering a question is input to training for obtaining step 3), obtain literary
This characteristic vector, then the characteristic vector of text is input in answer pond, obtain the characteristic vector of some candidate answers, Ran Houcong
The characteristic vector of each candidate answers chooses optimum answer, and exports the optimum answer, completes the intelligence doctor based on deep learning
Treat automatic question answering.
Pass through web crawlers technical limit spacing medical treatment question and answer data in step 2) from internet.
The expression formula of model training cost function is:
L=max { 0, M-s (q, a-)+s (q, a+)}
Wherein, q be problem text vector, a+The text vector of model answer, a are corresponded to for the problem-It is corresponding for the problem
The text vector of wrong answer, s (q, a+) it is q and a+Cosine similarity amount, s (q, a-) it is q and a-Cosine similarity amount, M
To be spaced constant.
The concrete operations for choosing optimum answer in step 4) from the characteristic vectors of each candidate answers are:Calculate the feature of text
The vectorial and cosine similarity of the characteristic vector of each candidate answers, then choose corresponding candidate answers during cosine similarity highest and make
For optimum answer.
The invention has the advantages that:
Intelligent medical automatic question-answering method of the present invention based on deep learning in concrete operations, deposit first by structure
Contain the answer pond of medical care problem, the answer of the medical care problem and the medical care problem art, then by magnanimity from interconnection
The medical question and answer data obtained on the net are input in neutral net as training text, and neutral net is trained, training
Neutral net afterwards can search candidate answers according to waiting to answer a question from answer pond, then be chosen from candidate answers optimal
Answer, and export, so as to realize the automatic answer of medical care problem, it is necessary to which explanation, the present invention is by magnanimity from internet
The medical question and answer data of acquisition are input in neutral net as training text, are realized and are trained study to the depth of neutral net,
So as to effectively improve question and answer accuracy, while, it is necessary to explanation, the present invention to each data carry out vectorization, pass through to
The mode of amount realizes the training of neutral net and the extraction of candidate answers, so as to effectively improve the accurate of candidate answers inquiry
Property, relative to traditional Keywords matching mode, the accuracy of flexibility, search efficiency and inquiry is higher.In practical application, this
The problem of invention is inputted by user is searched optimum answer automatically from answer pond and replied, and user is had science for the state of an illness
Rational anticipation, and rational choice medical treatment place, while effectively alleviate physician-patient relationship tense, the problem of medical resource is rare.
Brief description of the drawings
Fig. 1 is the schematic diagram of neutral net in the present invention;
Structural representation when Fig. 2 is training neutral net;
Fig. 3 is the schematic diagram of the present invention.
Embodiment
The present invention is described in further detail below in conjunction with the accompanying drawings:
With reference to figure 1, the intelligent medical automatic question-answering method of the present invention based on deep learning comprises the following steps:
1) structure is stored with the answer pond of medical care problem, the answer of the medical care problem and the medical care problem art,
Vectorization is carried out to the data in answer pond again;
It should be noted that the foundation in typical problem-answer storehouse and answer pond:For existing on-line communities question and answer data,
Mess code data are rejected and without answer data by the network information technology;Obtain and most preferably answered by what user chose under every problem
Case, then the medical question and answer data of structuring are made, the medical question and answer data of described structuring are by problem, answer and corresponding
Association area forms, and then builds answer pond by problem, answer and corresponding association area.
In addition, it is necessary to sequence number-text bilingual dictionary is built, specifically, jieba points provided first by python language
Word storehouse of increasing income is segmented to problem and answer, then the participle data of acquisition are numbered and build dictionary, is numbered with idx_+
Numeral is formed, and right side is word corresponding to numbering, wherein, to ensure each word in dictionary comprising only during dictionary is made
Word in one numbering and dictionary does not repeat.
To the disaggregated classification in answer pond, specifically, by using the association area in structured medical question and answer data to answer
Class is finely divided, can improve the degree of accuracy of follow-up answer matches, answer is divided into ten classes by the present invention, specifically includes internal medicine, outer
Section, paediatrics, gynemetrics, dept. of dermatology, ENT dept., andrology, contagious department, neurology department and department of traditional Chinese medicine.
It should be noted that the basic thought that distributed term vector characterizes is by training a word in certain language
The short amount of a regular length is mapped to, all short amounts are built into a term vector space, wherein, each vector is considered as
A point in term vector space, in this term vector space introduce distance concept, you can according between word and word away from
From the semantic and phraseological similitude judged between two words.
The distributed term vector that the present invention obtains question and answer text with text depth representing text depth representing model characterizes.
Text depth representing model is a efficient tool that word is characterized as to real number value vector, and it utilizes the thought of deep learning, can
The processing to content of text is reduced to by training the vector operation in K gts, and it is similar in vector space
Degree is used for representing the similarity on text semantic.Word is tieed up Feature Mapping to K using Word2vec as feature in the present invention
In vector space, you can judge the semantic and phraseological similitude between two words according to the distance between word and word.
Such as:Assuming that represent the K dimensional vectors of word " systolic pressure " as a1, the K dimensional vectors of word " diastolic pressure " are represented as a2,
The K dimensional vectors of word " great three positive " are represented as a3, from logic of language, " systolic pressure " and the semantic logic of " diastolic pressure " more connect
Closely, thus by text depth representing model by training obtained by vectorial a1 and a2 distance be less than a1 and a3 away from
From.
2) medical question and answer data are obtained from internet, then the medical question and answer data to acquiring carry out vectorization, obtain
The training text vector, is then input in neutral net by training text vector, and builds model training cost function, then
The neutral net is trained by model training cost function, the neutral net after must training;
3) treat to answer a question and carry out vectorization, then extract wait to answer a question in each word term vector, then will treat
The term vector of each word is spliced in answering a question, and must wait corresponding text vector of answering a question;
Specifically, being segmented first to problem, and the term vector of each word is matched, term vector is spliced one
Rise, form the sign matrix of problem, then using the sign matrix of problem as the input of neutral net, such as:Assuming that problem is
" diabetes are checked, I should be what if", problem is segmented, obtain " check/gone out/diabetes/,/I/should/
What if/", there is the replacement of K dimensional vectors to each word or punctuate in problem, then problem can tie up matrix by K*8 and replace;K*8 is tieed up
Input matrix obtains the hiding information of problem context into Recognition with Recurrent Neural Network, then by the hidden of Recognition with Recurrent Neural Network output
It is input to containing information in convolutional neural networks, finally carries out pondization operation again, obtains the feature of text.
4) in the neutral net after text vector corresponding to answering a question is input to training for obtaining step 3), obtain literary
This characteristic vector, then the characteristic vector of text is input in answer pond, obtain the characteristic vector of some candidate answers, Ran Hougen
Optimum answer is obtained according to the characteristic vector of each candidate answers, and exports the optimum answer, completes the intelligence based on deep learning
Medical automatic question answering.
Pass through web crawlers technical limit spacing medical treatment question and answer data in step 2) from internet.
The expression formula of model training cost function is:
L=max { 0, M-s (q, a-)+s (q, a+)}
Wherein, q be problem text vector, a+The text vector of model answer, a are corresponded to for the problem-It is corresponding for the problem
The text vector of wrong answer, s (q, a+) it is q and a+Cosine similarity amount, s (q, a-) it is q and a-Cosine similarity amount, M
To be spaced constant.
It is according to the concrete operations of the characteristic vector of each candidate answers acquisition optimum answer in step 4):Calculate the spy of text
The vectorial and characteristic vector of each candidate answers cosine similarity is levied, then chooses corresponding candidate answers during cosine similarity highest
As optimum answer.
Claims (4)
1. a kind of intelligent medical automatic question-answering method based on deep learning, it is characterised in that comprise the following steps:
1) structure is stored with the answer pond of medical care problem, the answer of the medical care problem and the medical care problem art, then right
Data in the answer pond carry out vectorization;
2) medical question and answer data are obtained from internet, then the medical question and answer data to acquiring carry out vectorization, must train
Text vector, then the training text vector is input in neutral net, and builds model training cost function, then passed through
Model training cost function is trained to the neutral net, the neutral net after must training;
3) treat to answer a question and carry out vectorization, then extract wait to answer a question in each word term vector, then will wait to answer
The term vector of each word is spliced in problem, must wait corresponding text vector of answering a question;
4) in the neutral net after text vector corresponding to answering a question is input to training for obtaining step 3), text is obtained
Characteristic vector, then the characteristic vector of text is input in answer pond, the characteristic vector of some candidate answers is obtained, then from each time
Select the characteristic vector of answer to choose optimum answer, and export the optimum answer, complete the intelligent medical based on deep learning certainly
Dynamic question and answer.
2. the intelligent medical automatic question-answering method according to claim 1 based on deep learning, it is characterised in that step 2)
In pass through web crawlers technical limit spacing medical treatment question and answer data from internet.
3. the intelligent medical automatic question-answering method according to claim 1 based on deep learning, it is characterised in that model is instructed
Practice cost function expression formula be:
L=max { 0, M-s (q, a-)+s (q, a+)}
Wherein, q be problem text vector, a+The text vector of model answer, a are corresponded to for the problem-Mistake is corresponded to for the problem
The text vector of answer, s (q, a+) it is q and a+Cosine similarity amount, s (q, a-) it is q and a-Cosine similarity amount, between M is
Every constant.
4. the intelligent medical automatic question-answering method according to claim 1 based on deep learning, it is characterised in that step 4)
In the concrete operations of optimum answer chosen from the characteristic vectors of each candidate answers be:The characteristic vector for calculating text is answered with each candidate
The cosine similarity of the characteristic vector of case, then corresponding candidate answers are chosen during cosine similarity highest as optimum answer.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710510408.7A CN107368547A (en) | 2017-06-28 | 2017-06-28 | A kind of intelligent medical automatic question-answering method based on deep learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710510408.7A CN107368547A (en) | 2017-06-28 | 2017-06-28 | A kind of intelligent medical automatic question-answering method based on deep learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107368547A true CN107368547A (en) | 2017-11-21 |
Family
ID=60305043
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710510408.7A Pending CN107368547A (en) | 2017-06-28 | 2017-06-28 | A kind of intelligent medical automatic question-answering method based on deep learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107368547A (en) |
Cited By (44)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108052588A (en) * | 2017-12-11 | 2018-05-18 | 浙江大学城市学院 | A kind of construction method of the document automatically request-answering system based on convolutional neural networks |
CN108090218A (en) * | 2017-12-29 | 2018-05-29 | 北京百度网讯科技有限公司 | Conversational system generation method and device based on deeply study |
CN108170853A (en) * | 2018-01-19 | 2018-06-15 | 广东惠禾科技发展有限公司 | A kind of chat language material method for self-cleaning, device and user terminal |
CN108231146A (en) * | 2017-12-01 | 2018-06-29 | 华南师范大学 | A kind of medical records model building method, system and device based on deep learning |
CN108256056A (en) * | 2018-01-12 | 2018-07-06 | 广州杰赛科技股份有限公司 | Intelligent answer method and system |
CN108345585A (en) * | 2018-01-11 | 2018-07-31 | 浙江大学 | A kind of automatic question-answering method based on deep learning |
CN108446321A (en) * | 2018-02-10 | 2018-08-24 | 武汉市灯塔互动文化传播有限公司 | A kind of automatic question-answering method based on deep learning |
CN108509539A (en) * | 2018-03-16 | 2018-09-07 | 联想(北京)有限公司 | Information processing method electronic equipment |
CN108509519A (en) * | 2018-03-09 | 2018-09-07 | 北京邮电大学 | World knowledge collection of illustrative plates enhancing question and answer interactive system based on deep learning and method |
CN108806785A (en) * | 2018-05-29 | 2018-11-13 | 四川长虹电器股份有限公司 | A kind of diagnosis and treatment section office recommendation method and system based on convolutional neural networks |
CN108846077A (en) * | 2018-06-08 | 2018-11-20 | 泰康保险集团股份有限公司 | Semantic matching method, device, medium and the electronic equipment of question and answer text |
CN108932349A (en) * | 2018-08-17 | 2018-12-04 | 齐鲁工业大学 | Medical automatic question-answering method and device, storage medium, electronic equipment |
CN108959404A (en) * | 2018-06-06 | 2018-12-07 | 北京春雨天下软件有限公司 | Intelligence questions closely method and device |
CN108960574A (en) * | 2018-06-07 | 2018-12-07 | 百度在线网络技术(北京)有限公司 | Quality determination method, device, server and the storage medium of question and answer |
CN109255016A (en) * | 2018-07-31 | 2019-01-22 | 平安科技(深圳)有限公司 | Answer method, device and computer readable storage medium based on deep learning |
CN109522393A (en) * | 2018-10-11 | 2019-03-26 | 平安科技(深圳)有限公司 | Intelligent answer method, apparatus, computer equipment and storage medium |
CN109543179A (en) * | 2018-11-05 | 2019-03-29 | 北京康夫子科技有限公司 | The normalized method and system of colloquial style symptom |
CN109684445A (en) * | 2018-11-13 | 2019-04-26 | 中国科学院自动化研究所 | Colloquial style medical treatment answering method and system |
CN109710772A (en) * | 2018-11-13 | 2019-05-03 | 国云科技股份有限公司 | A kind of question and answer library Knowledge Management System and its implementation based on deep learning |
CN109739958A (en) * | 2018-11-22 | 2019-05-10 | 普强信息技术(北京)有限公司 | A kind of specification handbook answering method and system |
CN109741824A (en) * | 2018-12-21 | 2019-05-10 | 质直(上海)教育科技有限公司 | A kind of medical way of inquisition based on machine learning |
CN109829042A (en) * | 2018-12-28 | 2019-05-31 | 北京科技大学 | A kind of knowledge Q-A system and method based on Biomedical literature |
CN109947928A (en) * | 2019-04-16 | 2019-06-28 | 上海海事大学 | A kind of retrieval type artificial intelligence question and answer robot development approach |
CN109949927A (en) * | 2019-02-18 | 2019-06-28 | 四川拾智联兴科技有限公司 | A kind of intelligent diagnosing method and its system based on deep neural network |
CN110162604A (en) * | 2019-01-24 | 2019-08-23 | 腾讯科技(深圳)有限公司 | Sentence generation method, device, equipment and storage medium |
CN110176311A (en) * | 2019-05-17 | 2019-08-27 | 北京印刷学院 | A kind of automatic medical proposal recommending method and system based on confrontation neural network |
CN110322959A (en) * | 2019-05-24 | 2019-10-11 | 山东大学 | A kind of Knowledge based engineering depth medical care problem method for routing and system |
CN110377719A (en) * | 2019-07-25 | 2019-10-25 | 广东工业大学 | Medical answering method and device |
CN110600123A (en) * | 2019-08-29 | 2019-12-20 | 成都信息工程大学 | Clinical auxiliary diagnosis method for traditional Chinese medicine |
CN110647914A (en) * | 2019-08-14 | 2020-01-03 | 深圳壹账通智能科技有限公司 | Intelligent service level training method and device and computer readable storage medium |
CN110825852A (en) * | 2019-11-07 | 2020-02-21 | 四川长虹电器股份有限公司 | Long text-oriented semantic matching method and system |
CN110928471A (en) * | 2018-09-19 | 2020-03-27 | 阿里健康信息技术有限公司 | Data display method and equipment terminal |
CN110968674A (en) * | 2019-12-04 | 2020-04-07 | 电子科技大学 | Method for constructing question comment pairs based on word vector representation |
CN111107437A (en) * | 2019-12-27 | 2020-05-05 | 深圳Tcl新技术有限公司 | Interaction method and system for movie and television after-viewing feeling, display terminal and readable storage medium |
CN111209385A (en) * | 2020-01-14 | 2020-05-29 | 重庆兆光科技股份有限公司 | Consultation session unique answer optimizing method based on convex neural network |
CN111309875A (en) * | 2018-12-10 | 2020-06-19 | 百度在线网络技术(北京)有限公司 | Question answering method, device, equipment and storage medium |
CN111368040A (en) * | 2018-12-25 | 2020-07-03 | 马上消费金融股份有限公司 | Dialogue processing method, model training method and related equipment |
CN111783473A (en) * | 2020-07-14 | 2020-10-16 | 腾讯科技(深圳)有限公司 | Method and device for identifying best answer in medical question and answer and computer equipment |
CN112116840A (en) * | 2019-06-19 | 2020-12-22 | 广东小天才科技有限公司 | Job correction method and system based on image recognition and intelligent terminal |
CN112164262A (en) * | 2020-11-09 | 2021-01-01 | 河南环球优路教育科技有限公司 | Intelligent paper reading tutoring system |
CN113012822A (en) * | 2021-03-23 | 2021-06-22 | 同济大学 | Medical question-answering system based on generating type dialogue technology |
CN113010655A (en) * | 2021-03-18 | 2021-06-22 | 华南理工大学 | Answer and interference item generation method and device for reading and understanding of machine |
CN114417880A (en) * | 2021-12-31 | 2022-04-29 | 福建亿榕信息技术有限公司 | Interactive intelligent question-answering method based on power grid practical training question-answering knowledge base |
US11568997B2 (en) | 2019-07-23 | 2023-01-31 | International Business Machines Corporation | Dynamic context-based collaborative medical concept interpreter |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160232444A1 (en) * | 2015-02-05 | 2016-08-11 | International Business Machines Corporation | Scoring type coercion for question answering |
CN106503123A (en) * | 2016-10-20 | 2017-03-15 | 宁波江东大金佰汇信息技术有限公司 | A kind of deep learning intelligent response system based on computer cloud data |
CN106599215A (en) * | 2016-12-16 | 2017-04-26 | 广州索答信息科技有限公司 | Question generation method and question generation system based on deep learning |
-
2017
- 2017-06-28 CN CN201710510408.7A patent/CN107368547A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20160232444A1 (en) * | 2015-02-05 | 2016-08-11 | International Business Machines Corporation | Scoring type coercion for question answering |
CN106503123A (en) * | 2016-10-20 | 2017-03-15 | 宁波江东大金佰汇信息技术有限公司 | A kind of deep learning intelligent response system based on computer cloud data |
CN106599215A (en) * | 2016-12-16 | 2017-04-26 | 广州索答信息科技有限公司 | Question generation method and question generation system based on deep learning |
Non-Patent Citations (1)
Title |
---|
BAOTIAN HU 等: "Convolutional Neural Network Architectures for Matching Natural Language Sentences", 《PROCEEDING NIPS"14 PROCEEDINGS OF THE 27TH INTERNATIONAL CONFERENCE ON NEURAL INFORMATION PROCESSING SYSTEMS》 * |
Cited By (64)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108231146A (en) * | 2017-12-01 | 2018-06-29 | 华南师范大学 | A kind of medical records model building method, system and device based on deep learning |
CN108052588A (en) * | 2017-12-11 | 2018-05-18 | 浙江大学城市学院 | A kind of construction method of the document automatically request-answering system based on convolutional neural networks |
CN108052588B (en) * | 2017-12-11 | 2021-03-26 | 浙江大学城市学院 | Method for constructing automatic document question-answering system based on convolutional neural network |
CN108090218A (en) * | 2017-12-29 | 2018-05-29 | 北京百度网讯科技有限公司 | Conversational system generation method and device based on deeply study |
CN108090218B (en) * | 2017-12-29 | 2022-08-23 | 北京百度网讯科技有限公司 | Dialog system generation method and device based on deep reinforcement learning |
CN108345585A (en) * | 2018-01-11 | 2018-07-31 | 浙江大学 | A kind of automatic question-answering method based on deep learning |
CN108256056A (en) * | 2018-01-12 | 2018-07-06 | 广州杰赛科技股份有限公司 | Intelligent answer method and system |
CN108170853A (en) * | 2018-01-19 | 2018-06-15 | 广东惠禾科技发展有限公司 | A kind of chat language material method for self-cleaning, device and user terminal |
CN108170853B (en) * | 2018-01-19 | 2020-06-19 | 广东惠禾科技发展有限公司 | Chat corpus self-cleaning method and device and user terminal |
CN108446321A (en) * | 2018-02-10 | 2018-08-24 | 武汉市灯塔互动文化传播有限公司 | A kind of automatic question-answering method based on deep learning |
CN108509519B (en) * | 2018-03-09 | 2021-03-09 | 北京邮电大学 | General knowledge graph enhanced question-answer interaction system and method based on deep learning |
CN108509519A (en) * | 2018-03-09 | 2018-09-07 | 北京邮电大学 | World knowledge collection of illustrative plates enhancing question and answer interactive system based on deep learning and method |
CN108509539B (en) * | 2018-03-16 | 2021-08-17 | 联想(北京)有限公司 | Information processing method and electronic device |
CN108509539A (en) * | 2018-03-16 | 2018-09-07 | 联想(北京)有限公司 | Information processing method electronic equipment |
CN108806785A (en) * | 2018-05-29 | 2018-11-13 | 四川长虹电器股份有限公司 | A kind of diagnosis and treatment section office recommendation method and system based on convolutional neural networks |
CN108959404A (en) * | 2018-06-06 | 2018-12-07 | 北京春雨天下软件有限公司 | Intelligence questions closely method and device |
CN108960574A (en) * | 2018-06-07 | 2018-12-07 | 百度在线网络技术(北京)有限公司 | Quality determination method, device, server and the storage medium of question and answer |
CN108846077A (en) * | 2018-06-08 | 2018-11-20 | 泰康保险集团股份有限公司 | Semantic matching method, device, medium and the electronic equipment of question and answer text |
WO2020024452A1 (en) * | 2018-07-31 | 2020-02-06 | 平安科技(深圳)有限公司 | Deep learning-based answering method and apparatus, and readable storage medium |
CN109255016A (en) * | 2018-07-31 | 2019-01-22 | 平安科技(深圳)有限公司 | Answer method, device and computer readable storage medium based on deep learning |
CN108932349A (en) * | 2018-08-17 | 2018-12-04 | 齐鲁工业大学 | Medical automatic question-answering method and device, storage medium, electronic equipment |
AU2019322953B2 (en) * | 2018-08-17 | 2021-08-19 | Qilu University Of Technology | Method, system, storage medium and electric device of medical automatic question answering |
CN108932349B (en) * | 2018-08-17 | 2019-03-26 | 齐鲁工业大学 | Medical automatic question-answering method and device, storage medium, electronic equipment |
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 |
CN110928471B (en) * | 2018-09-19 | 2023-12-26 | 阿里健康信息技术有限公司 | Data display method and equipment terminal |
CN110928471A (en) * | 2018-09-19 | 2020-03-27 | 阿里健康信息技术有限公司 | Data display method and equipment terminal |
CN109522393A (en) * | 2018-10-11 | 2019-03-26 | 平安科技(深圳)有限公司 | Intelligent answer method, apparatus, computer equipment and storage medium |
CN109543179A (en) * | 2018-11-05 | 2019-03-29 | 北京康夫子科技有限公司 | The normalized method and system of colloquial style symptom |
CN109684445A (en) * | 2018-11-13 | 2019-04-26 | 中国科学院自动化研究所 | Colloquial style medical treatment answering method and system |
CN109710772A (en) * | 2018-11-13 | 2019-05-03 | 国云科技股份有限公司 | A kind of question and answer library Knowledge Management System and its implementation based on deep learning |
CN109739958A (en) * | 2018-11-22 | 2019-05-10 | 普强信息技术(北京)有限公司 | A kind of specification handbook answering method and system |
CN111309875A (en) * | 2018-12-10 | 2020-06-19 | 百度在线网络技术(北京)有限公司 | Question answering method, device, equipment and storage medium |
CN111309875B (en) * | 2018-12-10 | 2023-08-04 | 百度在线网络技术(北京)有限公司 | Method, device, equipment and storage medium for answering questions |
CN109741824B (en) * | 2018-12-21 | 2023-08-04 | 质直(上海)教育科技有限公司 | Medical inquiry method based on machine learning |
CN109741824A (en) * | 2018-12-21 | 2019-05-10 | 质直(上海)教育科技有限公司 | A kind of medical way of inquisition based on machine learning |
CN111368040B (en) * | 2018-12-25 | 2021-01-26 | 马上消费金融股份有限公司 | Dialogue processing method, model training method and related equipment |
CN111368040A (en) * | 2018-12-25 | 2020-07-03 | 马上消费金融股份有限公司 | Dialogue processing method, model training method and related equipment |
CN109829042A (en) * | 2018-12-28 | 2019-05-31 | 北京科技大学 | A kind of knowledge Q-A system and method based on Biomedical literature |
CN110162604A (en) * | 2019-01-24 | 2019-08-23 | 腾讯科技(深圳)有限公司 | Sentence generation method, device, equipment and storage medium |
CN110162604B (en) * | 2019-01-24 | 2023-09-12 | 腾讯科技(深圳)有限公司 | Statement generation method, device, equipment and storage medium |
WO2020151690A1 (en) * | 2019-01-24 | 2020-07-30 | 腾讯科技(深圳)有限公司 | Statement generation method, device and equipment and storage medium |
CN109949927A (en) * | 2019-02-18 | 2019-06-28 | 四川拾智联兴科技有限公司 | A kind of intelligent diagnosing method and its system based on deep neural network |
CN109947928A (en) * | 2019-04-16 | 2019-06-28 | 上海海事大学 | A kind of retrieval type artificial intelligence question and answer robot development approach |
CN110176311A (en) * | 2019-05-17 | 2019-08-27 | 北京印刷学院 | A kind of automatic medical proposal recommending method and system based on confrontation neural network |
CN110322959A (en) * | 2019-05-24 | 2019-10-11 | 山东大学 | A kind of Knowledge based engineering depth medical care problem method for routing and system |
CN110322959B (en) * | 2019-05-24 | 2021-09-28 | 山东大学 | Deep medical problem routing method and system based on knowledge |
CN112116840A (en) * | 2019-06-19 | 2020-12-22 | 广东小天才科技有限公司 | Job correction method and system based on image recognition and intelligent terminal |
US11568997B2 (en) | 2019-07-23 | 2023-01-31 | International Business Machines Corporation | Dynamic context-based collaborative medical concept interpreter |
CN110377719A (en) * | 2019-07-25 | 2019-10-25 | 广东工业大学 | Medical answering method and device |
CN110647914A (en) * | 2019-08-14 | 2020-01-03 | 深圳壹账通智能科技有限公司 | Intelligent service level training method and device and computer readable storage medium |
CN110600123A (en) * | 2019-08-29 | 2019-12-20 | 成都信息工程大学 | Clinical auxiliary diagnosis method for traditional Chinese medicine |
CN110825852A (en) * | 2019-11-07 | 2020-02-21 | 四川长虹电器股份有限公司 | Long text-oriented semantic matching method and system |
CN110968674B (en) * | 2019-12-04 | 2023-04-18 | 电子科技大学 | Method for constructing question and comment pairs based on word vector representation |
CN110968674A (en) * | 2019-12-04 | 2020-04-07 | 电子科技大学 | Method for constructing question comment pairs based on word vector representation |
CN111107437A (en) * | 2019-12-27 | 2020-05-05 | 深圳Tcl新技术有限公司 | Interaction method and system for movie and television after-viewing feeling, display terminal and readable storage medium |
CN111209385A (en) * | 2020-01-14 | 2020-05-29 | 重庆兆光科技股份有限公司 | Consultation session unique answer optimizing method based on convex neural network |
CN111209385B (en) * | 2020-01-14 | 2024-02-02 | 重庆兆光科技股份有限公司 | Convex neural network-based consultation dialogue unique answer optimizing method |
CN111783473A (en) * | 2020-07-14 | 2020-10-16 | 腾讯科技(深圳)有限公司 | Method and device for identifying best answer in medical question and answer and computer equipment |
CN111783473B (en) * | 2020-07-14 | 2024-02-13 | 腾讯科技(深圳)有限公司 | Method and device for identifying best answer in medical question and answer and computer equipment |
CN112164262A (en) * | 2020-11-09 | 2021-01-01 | 河南环球优路教育科技有限公司 | Intelligent paper reading tutoring system |
CN113010655A (en) * | 2021-03-18 | 2021-06-22 | 华南理工大学 | Answer and interference item generation method and device for reading and understanding of machine |
CN113010655B (en) * | 2021-03-18 | 2022-12-16 | 华南理工大学 | Answer and interference item generation method and device for reading and understanding of machine |
CN113012822A (en) * | 2021-03-23 | 2021-06-22 | 同济大学 | Medical question-answering system based on generating type dialogue technology |
CN114417880A (en) * | 2021-12-31 | 2022-04-29 | 福建亿榕信息技术有限公司 | Interactive intelligent question-answering method based on power grid practical training question-answering knowledge base |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107368547A (en) | A kind of intelligent medical automatic question-answering method based on deep learning | |
CN107391906B (en) | Healthy diet knowledge network construction method based on neural network and map structure | |
CN110222201B (en) | Method and device for constructing special disease knowledge graph | |
CN110032648A (en) | A kind of case history structuring analytic method based on medical domain entity | |
CN112487202B (en) | Chinese medical named entity recognition method and device fusing knowledge map and BERT | |
CN106776711A (en) | A kind of Chinese medical knowledge mapping construction method based on deep learning | |
CN108538395A (en) | A kind of construction method of general medical disease that calls for specialized treatment data system | |
CN112786194A (en) | Medical image diagnosis guide inspection system, method and equipment based on artificial intelligence | |
CN110598000A (en) | Relationship extraction and knowledge graph construction method based on deep learning model | |
CN108628824A (en) | A kind of entity recognition method based on Chinese electronic health record | |
CN106202955A (en) | Diagnosis associated packets method and system based on intellectual coded adaptation | |
CN112148851A (en) | Construction method of medicine knowledge question-answering system based on knowledge graph | |
CN110838368A (en) | Robot active inquiry method based on traditional Chinese medicine clinical knowledge graph | |
Chen et al. | A benchmark for automatic medical consultation system: frameworks, tasks and datasets | |
CN111259111B (en) | Medical record-based decision-making assisting method and device, electronic equipment and storage medium | |
CN101149751A (en) | Generalized relating rule digging method for analyzing traditional Chinese medicine recipe drug matching rule | |
CN116805013A (en) | Traditional Chinese medicine video retrieval model based on knowledge graph | |
CN115293161A (en) | Reasonable medicine taking system and method based on natural language processing and medicine knowledge graph | |
CN114676260A (en) | Human body bone motion rehabilitation model construction method based on knowledge graph | |
CN117171329A (en) | Semantic analysis-based traditional Chinese medicine domain knowledge graph question-answering method | |
CN112883172A (en) | Biomedical question-answering method based on dual knowledge selection | |
CN116702743A (en) | Text similarity detection method and device, electronic equipment and storage medium | |
Mucheng et al. | TCM-SD: a benchmark for probing syndrome differentiation via Natural Language processing | |
CN115658863A (en) | Question-answering system construction method based on diabetes knowledge graph | |
CN113314236A (en) | Intelligent question-answering system for hypertension |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171121 |