CN109657047B - Voice automatic question-answering method and system based on crawler technology and machine learning - Google Patents
Voice automatic question-answering method and system based on crawler technology and machine learning Download PDFInfo
- Publication number
- CN109657047B CN109657047B CN201811619723.4A CN201811619723A CN109657047B CN 109657047 B CN109657047 B CN 109657047B CN 201811619723 A CN201811619723 A CN 201811619723A CN 109657047 B CN109657047 B CN 109657047B
- Authority
- CN
- China
- Prior art keywords
- answers
- final answer
- lifting tree
- question
- internet
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
- G10L2015/225—Feedback of the input speech
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Machine Translation (AREA)
Abstract
The invention discloses a voice automatic question-answering method and system based on crawler technology and machine learning, and belongs to the technical field of voice recognition, machine learning and web crawler. Analyzing the obtained questioning voice to obtain a question keyword set; based on a lifting tree algorithm, filtering and retrieving the problem keyword set, and then obtaining a final answer set from the result after filtering and retrieving; and selecting a final answer from the final answer set, processing the filtered and retrieved result according to the final answer, and filtering the retrieved result after processing for next selection. The invention is used for automatic voice question answering.
Description
Technical Field
A voice automatic question-answering method and system based on crawler technology and machine learning are used for voice automatic question-answering and belong to the technical field of voice recognition, machine learning and web crawler.
Background
Microsoft Speech recognition module (Microsoft Speech SDK): microsoft Speech SDK is a software development kit provided by Microsoft, and the provided Speech API (SAPI) mainly comprises two main aspects:
1.API for Text-to-Speech
2.API for Speech Recognition
the API for Text-to-Speech is the interface of the Microsoft TTS engine, and a Text voice program with powerful functions can be easily established through the API, the writing API is used for the word reading function of the Jinshan worship, and almost all the current Text reading tools are developed by the SDK. As for the API for Speech Recognition corresponding to TTS, the Speech technology is an exciting technology, but the accuracy and Recognition speed of the current Speech Recognition technology are not ideal, and the requirements of wide application are not met.
Stanford dependent syntactic analysis (Stanford CoreNLP): stanford parser is an open source parser developed by the Stanford university Natural language processing group, a JAVA implementation based on probabilistic statistical parsing. The parser currently provides 5 chinese grammars.
Web crawlers: a web crawler (also known as a web spider, web robot, among FOAF communities, and more often referred to as a web chaser) is a program or script that automatically crawls the world wide web according to certain rules.
Machine learning: it is the core of artificial intelligence, and is a fundamental way for computer to possess intelligence, and its application is extensive in every field of artificial intelligence, and it mainly uses induction, synthesis, rather than deduction.
And (3) a lifting tree algorithm: boosting is a widely used and very effective statistical learning method. It is based on the idea that: for a complex task, the judgment obtained by properly combining the judgment of multiple experts is better than the judgment of any one expert alone.
In the existing automatic answering system, a user generally inputs a question keyword, the answering system retrieves a corresponding question from a question bank according to the keyword, and then the answer is displayed to the user.
Yet another way is to use speech recognition technology, where the user can ask questions by speech. The answer to the question is still to manual entry.
The answer of the two modes depends on manual input, one part is system management personnel, and the other part is from active input of a user, so that the following defects exist:
firstly, the input and the update of answers in the question bank depend on manual input of personnel, the accuracy of the answers requires that related personnel have related specialties, if the input is wrong, the accurate answers cannot be found, the input is very time-consuming, the waste of human resources can be caused, and the like;
secondly, the answers are limited and fixed, the latest answer of the question cannot be updated at any time, and the problem of poor user experience is caused.
Disclosure of Invention
Aiming at the problems of the research, the invention aims to provide an intelligent response method and an intelligent response system based on voice recognition and machine learning, and solves the problems that answers of intelligent response in the prior art are all dependent on manual input, input information is limited, accuracy of response information is low, updating is not timely, and professional requirements on input personnel are high.
In order to achieve the purpose, the invention adopts the following technical scheme:
an intelligent response method based on voice recognition and machine learning is characterized by comprising the following steps:
step 1, analyzing the obtained question voice to obtain a question keyword set;
step 2, based on a lifting tree algorithm, filtering and retrieving the problem keyword set, and then obtaining a final answer set from the result after filtering and retrieving;
and 3, selecting a final answer from the final answer set, processing the filtered and retrieved result according to the final answer, and replacing the filtered and retrieved result in the step 2 after processing for next selection.
Further, the specific steps of step 1 are:
step 1.1, identifying the questioning voice to obtain a question;
and step 1.2, decomposing the problem by using Stanford dependency syntax analysis to obtain a keyword set.
Further, the specific steps of step 2 are:
2.1, filtering and analyzing the keywords in the keyword set based on a grammar library, and removing subject and language help words to obtain a template problem;
2.2, reading and extending each keyword in the template problem;
step 2.3, searching and interpreting answers of the extended template questions in the question bank to obtain the existing basic answers and the weights corresponding to the existing basic answers, wherein the obtained existing basic answers and the weights corresponding to the existing basic answers are an existing lifting tree;
2.4, searching the extended template problem in the Internet by adopting a web crawler mode to obtain Internet basic answers and weights corresponding to the Internet basic answers, and establishing an Internet lifting tree according to the obtained result and a lifting tree algorithm;
and 2.5, combining the existing lifting tree in the step 2.3 and the Internet lifting tree in the step 2.4 to obtain a new lifting tree, and selecting basic answers with N bits before weight ranking from the new lifting tree to obtain a final answer set.
Further, the specific steps of step 3 are:
step 3.1, the user selects a final answer from the final answer set;
and 3.2, recording the final answer, pruning the new lifting tree with the added weight after the weight of the answer in the new lifting tree corresponding to the question is added, removing part of answers with too low weights, and replacing the new lifting tree obtained before pruning after the answer with too low weights is removed for next selection.
An intelligent response system based on speech recognition and machine learning, comprising:
a voice recognition module: analyzing the obtained questioning voice to obtain a question keyword set;
a processing module: based on a lifting tree algorithm, filtering and retrieving the problem keyword set, and then obtaining a final answer set from the results after filtering and retrieving or receiving feedback to replace the results after filtering and retrieving;
a selection processing module: and selecting a final answer from the final answer set, processing the filtered and retrieved result according to the final answer, and feeding back the processed result to the processing module.
Further, the implementation manner of the voice recognition module comprises the following steps:
identifying the questioning voice to obtain a question;
the problem is decomposed using Stenford dependency parsing to obtain a set of keywords.
Further, the implementation manner of the processing module comprises the following steps:
filtering and analyzing the keywords in the keyword set based on a grammar library, and removing subject and language help words to obtain a template problem;
reading and extending each keyword in the template problem;
searching and interpreting answers of the extended template questions in the question bank to obtain the existing basic answers and the weights corresponding to the existing basic answers, wherein the obtained existing basic answers and the weights corresponding to the existing basic answers are an existing lifting tree;
searching the extended template problem in the Internet by adopting a web crawler mode to obtain Internet basic answers and weights corresponding to the Internet answers, and establishing an Internet lifting tree according to the obtained result and a lifting tree algorithm;
and combining the two kinds of lifting trees to obtain a new lifting tree, and selecting the basic answers with the N-bit weights before ranking from the new lifting tree to obtain a final answer set.
Further, the implementation manner of the selection processing module includes the following steps:
the user selects a final answer from the final answer set;
recording the final answer, increasing the weight of the answer in the new lifting tree corresponding to the question, pruning the new lifting tree after increasing the weight, removing part of answers with too low weights, and feeding the result back to the processing module after removing the answers with too low weights.
Compared with the prior art, the invention has the beneficial effects that:
the invention adopts the web crawler technology to grab the latest answer of the questioning voice on the internet in real time, thereby avoiding the burden of manual input; meanwhile, a machine learning mode is adopted, so that the use of the user can be utilized while the answer which accords with the mind is selected for the user through the machine learning (the tree lifting algorithm), and the accuracy of solving the problem is improved;
the question bank is from the Internet and obtained by machine learning, the latest answer of the Internet can be updated at any time, and meanwhile, the action of selecting the answer by the user is also used for training AI (AI), so that the accuracy of the next selection is improved (namely the probability of the corresponding answer is improved);
thirdly, the combined new lifting tree is pruned, certain answers of non-sticky edges on the Internet can be removed, and redundant data are reduced;
and fourthly, the network crawler technology and the machine learning mode are adopted, so that the labor cost can be reduced.
Drawings
FIG. 1 is a schematic flow chart of the present invention.
Detailed Description
The invention will be further described with reference to the accompanying drawings and specific embodiments.
An intelligent response method based on voice recognition and machine learning comprises the following steps:
step 1, analyzing the obtained question voice to obtain a question keyword set; the method comprises the following specific steps:
step 1.1, identifying the questioning voice to obtain a question;
and step 1.2, decomposing the problem by using Stanford dependency syntax analysis to obtain a keyword set.
Step 2, based on a lifting tree algorithm, filtering and retrieving the problem keyword set, and then obtaining a final answer set from the result after filtering and retrieving; the method comprises the following specific steps:
2.1, filtering and analyzing the keywords in the keyword set based on a grammar library, and removing subject and language help words to obtain a template problem;
2.2, reading and extending each keyword in the template problem;
step 2.3, searching and interpreting answers of the extended template questions in the question bank to obtain the existing basic answers and the weights corresponding to the existing basic answers, wherein the obtained existing basic answers and the weights corresponding to the existing basic answers are an existing lifting tree;
2.4, searching the extended template problem in the Internet by adopting a web crawler mode to obtain Internet basic answers and weights corresponding to the Internet basic answers, and establishing an Internet lifting tree according to the obtained result and a lifting tree algorithm;
and 2.5, combining the existing promotion tree in the step 2.3 and the Internet promotion tree in the step 2.4 to obtain a new promotion tree (namely a result after filtering and searching), and selecting basic answers with N bits before weight ranking from the new promotion tree to obtain a final answer set.
And 3, selecting a final answer from the final answer set, processing the filtered and retrieved result according to the final answer, and replacing the filtered and retrieved result in the step 2 after processing for next selection. The method comprises the following specific steps:
step 3.1, the user selects a final answer from the final answer set;
and 3.2, recording the final answer, pruning the new lifting tree with the added weight after the weight of the answer in the new lifting tree corresponding to the question is added, removing part of answers with too low weights, and replacing the new lifting tree obtained before pruning after the answer with too low weights is removed for next selection.
An intelligent answering system based on speech recognition and machine learning, comprising:
a voice recognition module: analyzing the obtained questioning voice to obtain a question keyword set; the implementation mode comprises the following steps:
identifying the questioning voice to obtain a question;
the problem is decomposed using Stenford dependency parsing to obtain a set of keywords.
A processing module: based on a lifting tree algorithm, filtering and retrieving the problem keyword set, and then obtaining a final answer set from the results after filtering and retrieving or receiving feedback to replace the results after filtering and retrieving; the implementation mode comprises the following steps:
filtering and analyzing the keywords in the keyword set based on a grammar library, and removing subject and language help words to obtain a template problem;
reading and extending each keyword in the template problem;
searching and interpreting answers of the extended template questions in the question bank to obtain the existing basic answers and the weights corresponding to the existing basic answers, wherein the obtained existing basic answers and the weights corresponding to the existing basic answers are an existing lifting tree;
searching the extended template problem in the Internet by adopting a web crawler mode to obtain Internet basic answers and weights corresponding to the Internet answers, and establishing an Internet lifting tree according to the obtained result and a lifting tree algorithm;
and combining the two kinds of lifting trees to obtain a new lifting tree, and selecting the basic answers with the N-bit weights before ranking from the new lifting tree to obtain a final answer set.
A selection processing module: and selecting a final answer from the final answer set, processing the filtered and retrieved result according to the final answer, and feeding back the processed result to the processing module. The implementation mode comprises the following steps:
the user selects a final answer from the final answer set;
recording the final answer, increasing the weight of the answer in the new lifting tree corresponding to the question, pruning the new lifting tree after increasing the weight, removing part of answers with too low weights, and feeding the result back to the processing module after removing the answers with too low weights.
Examples
The recognition of Speech using the Microsoft Speech recognition module (Microsoft Speech SDK) results in problems, such as: "what I should eat today"; the problem is decomposed using Stanford CoreNLP (Stanford dependency syntax analysis) to get a set of keywords as I, today, supposed, eaten, and what.
The grammar library selects a Baidu natural language processing API, lexical analysis can be performed, the part of speech is recognized, filtering and analyzing are performed on the keywords in the keyword set based on the grammar library, namely, what I, today, supposed, eaten and what are analyzed, subject words, language help words and the like are eliminated, and what the template problem is eaten today is obtained; extending each keyword of the template problem, and reading various keywords such as time words, for example, what is eaten today, what is eaten on Saturday, twenty-thirteen days in December and winter solstice, can be obtained; searching answers in a question bank to obtain existing basic answers (leaf nodes), wherein if 23 apples are eaten today, 4 vegetables are eaten today, 35 meats are eaten today, 23, 4 and 35 are weights of the existing basic answers, the obtained existing basic answers and the weights corresponding to the existing basic answers are a lifting tree, the question bank is self-built and used for storing questions and the corresponding lifting tree, and therefore the retrieved questions are the lifting tree;
the method comprises the steps of searching questions on the internet (such as hectic question answering, known answer and the like) by adopting a web crawler mode to obtain internet basic answers, calculating weights according to the frequency of the internet basic answers, and establishing an internet promotion tree based on a promotion tree algorithm, the obtained internet basic answers and the weights corresponding to the internet basic answers, wherein for example, 1 porridge is eaten on saturday, 7 rice cakes are eaten on twenty-three months and 13 dumplings are eaten on winter; and merging the two promotion trees, and selecting the basic answers with weights of N bits before ranking from the new promotion trees to obtain a final answer set, wherein the final answer set comprises the steps that I eat apples today, I eat meat today, I eat rice cakes in twenty-three months and I eat dumplings in winter.
Selecting and recording a final answer in the final answer set by the user, and increasing the weight of the answer in the lifting tree corresponding to the question, for example, the user selects the Chinese winter solstice to eat the dumpling; increasing the weight of the answer in the lifting tree corresponding to the question of 'what I should eat today', pruning a new lifting tree after increasing the weight, removing part of answers with too low weights, training an AI (artificial intelligence) to prepare for the next selection of the user, and training A1, namely replacing the pruned new lifting tree with a new lifting tree.
The above are merely representative examples of the many specific applications of the present invention, and do not limit the scope of the invention in any way. All the technical solutions formed by the transformation or the equivalent substitution fall within the protection scope of the present invention.
Claims (6)
1. An intelligent response method based on voice recognition and machine learning is characterized by comprising the following steps:
step 1, analyzing the obtained question voice to obtain a question keyword set;
step 2, based on a lifting tree algorithm, filtering and retrieving the problem keyword set, and then obtaining a final answer set from the result after filtering and retrieving;
the specific steps of the step 2 are as follows:
2.1, filtering and analyzing the keywords in the keyword set based on a grammar library, and removing subject and language help words to obtain a template problem;
2.2, reading and extending each keyword in the template problem;
step 2.3, searching and interpreting answers of the extended template questions in the question bank to obtain the existing basic answers and the weights corresponding to the existing basic answers, wherein the obtained existing basic answers and the weights corresponding to the existing basic answers are an existing lifting tree;
2.4, searching the extended template problem in the Internet by adopting a web crawler mode to obtain Internet basic answers and weights corresponding to the Internet basic answers, and establishing an Internet lifting tree according to the obtained result and a lifting tree algorithm;
step 2.5, combining the existing lifting tree in the step 2.3 and the Internet lifting tree in the step 2.4 to obtain a new lifting tree, and selecting basic answers with N-bit weights before ranking from the new lifting tree to obtain a final answer set;
and 3, selecting a final answer from the final answer set, processing the filtered and retrieved result according to the final answer, and replacing the filtered and retrieved result in the step 2 after processing for next selection.
2. The intelligent response method based on speech recognition and machine learning according to claim 1, characterized in that the specific steps of step 1 are:
step 1.1, identifying the questioning voice to obtain a question;
and step 1.2, decomposing the problem by using Stanford dependency syntax analysis to obtain a keyword set.
3. The intelligent response method based on speech recognition and machine learning according to claim 1, characterized in that the specific steps of step 3 are:
step 3.1, the user selects a final answer from the final answer set;
and 3.2, recording the final answer, pruning the new lifting tree with the added weight after the weight of the answer in the new lifting tree corresponding to the question is added, removing part of answers with too low weights, and replacing the new lifting tree obtained before pruning after the answer with too low weights is removed for next selection.
4. An intelligent response system based on speech recognition and machine learning, comprising:
a voice recognition module: analyzing the obtained questioning voice to obtain a question keyword set;
a processing module: based on a lifting tree algorithm, filtering and retrieving the problem keyword set, and then obtaining a final answer set from the results after filtering and retrieving or receiving feedback to replace the results after filtering and retrieving;
the implementation mode of the processing module comprises the following steps:
filtering and analyzing the keywords in the keyword set based on a grammar library, and removing subject and language help words to obtain a template problem;
reading and extending each keyword in the template problem;
searching and interpreting answers of the extended template questions in the question bank to obtain the existing basic answers and the weights corresponding to the existing basic answers, wherein the obtained existing basic answers and the weights corresponding to the existing basic answers are an existing lifting tree;
searching the extended template problem in the Internet by adopting a web crawler mode to obtain Internet basic answers and weights corresponding to the Internet basic answers, and establishing an Internet lifting tree according to the obtained result and a lifting tree algorithm;
combining the existing promotion tree and the Internet promotion tree to obtain a new promotion tree, and selecting basic answers with N bits before weight ranking from the new promotion tree to obtain a final answer set;
a selection processing module: and selecting a final answer from the final answer set, processing the filtered and retrieved result according to the final answer, and feeding back the processed result to the processing module.
5. The intelligent response system based on speech recognition and machine learning of claim 4, wherein the implementation manner of the speech recognition module comprises the following steps:
identifying the questioning voice to obtain a question;
the problem is decomposed using Stenford dependency parsing to obtain a set of keywords.
6. The intelligent response system based on speech recognition and machine learning of claim 4, wherein the implementation manner of the selection processing module comprises the following steps:
the user selects a final answer from the final answer set;
recording the final answer, increasing the weight of the answer in the new lifting tree corresponding to the question, pruning the new lifting tree after increasing the weight, removing part of answers with too low weights, and feeding the result back to the processing module after removing the answers with too low weights.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811619723.4A CN109657047B (en) | 2018-12-27 | 2018-12-27 | Voice automatic question-answering method and system based on crawler technology and machine learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811619723.4A CN109657047B (en) | 2018-12-27 | 2018-12-27 | Voice automatic question-answering method and system based on crawler technology and machine learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109657047A CN109657047A (en) | 2019-04-19 |
| CN109657047B true CN109657047B (en) | 2020-09-29 |
Family
ID=66117547
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811619723.4A Active CN109657047B (en) | 2018-12-27 | 2018-12-27 | Voice automatic question-answering method and system based on crawler technology and machine learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109657047B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111782794A (en) * | 2020-05-29 | 2020-10-16 | 北京沃东天骏信息技术有限公司 | Question-answer response method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107220380A (en) * | 2017-06-27 | 2017-09-29 | 北京百度网讯科技有限公司 | Question and answer based on artificial intelligence recommend method, device and computer equipment |
| CN108595696A (en) * | 2018-05-09 | 2018-09-28 | 长沙学院 | A kind of human-computer interaction intelligent answering method and system based on cloud platform |
| CN108717433A (en) * | 2018-05-14 | 2018-10-30 | 南京邮电大学 | A kind of construction of knowledge base method and device of programming-oriented field question answering system |
| 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 |
-
2018
- 2018-12-27 CN CN201811619723.4A patent/CN109657047B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107220380A (en) * | 2017-06-27 | 2017-09-29 | 北京百度网讯科技有限公司 | Question and answer based on artificial intelligence recommend method, device and computer equipment |
| CN108595696A (en) * | 2018-05-09 | 2018-09-28 | 长沙学院 | A kind of human-computer interaction intelligent answering method and system based on cloud platform |
| CN108717433A (en) * | 2018-05-14 | 2018-10-30 | 南京邮电大学 | A kind of construction of knowledge base method and device of programming-oriented field question answering system |
| 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 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109657047A (en) | 2019-04-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110337645B (en) | Adaptable processing assembly | |
| CN111143576A (en) | Event-oriented dynamic knowledge graph construction method and device | |
| CN110110041A (en) | Wrong word correcting method, device, computer installation and storage medium | |
| CN106649768A (en) | Deep question answering-based questions and answers clarifying method and device | |
| WO2020010834A1 (en) | Faq question and answer library generalization method, apparatus, and device | |
| CN111324721A (en) | Method for constructing intelligent question-answering knowledge base | |
| CN111930948B (en) | Information collection and classification method and device, computer equipment and storage medium | |
| CN107832439A (en) | Method, system and the terminal device of more wheel state trackings | |
| CN110851584B (en) | Legal provision accurate recommendation system and method | |
| CN116166785A (en) | Multi-intention recognition method, equipment and storage medium based on event extraction | |
| CN109657047B (en) | Voice automatic question-answering method and system based on crawler technology and machine learning | |
| CN111881264B (en) | Method and electronic equipment for searching long text in question-answering task in open field | |
| CN114021546A (en) | Peach production knowledge open question-answering method and device for migrating context network | |
| CN110377706B (en) | Search sentence mining method and device based on deep learning | |
| CN112200674A (en) | Stock market emotion index intelligent calculation information system | |
| CN111522913A (en) | Emotion classification method suitable for long text and short text | |
| CN107368464B (en) | Method and device for acquiring bidding product information | |
| CN115017271A (en) | Method and system for intelligently generating RPA flow component block | |
| Shahbaz et al. | Automatic generation of extended er diagram using natural language processing | |
| CN114970733A (en) | Corpus generation method, apparatus, system, storage medium and electronic device | |
| WO2023098971A1 (en) | Method and apparatus for self-supervised extractive question answering | |
| CN114417008A (en) | Construction engineering field-oriented knowledge graph construction method and system | |
| CN114020774A (en) | Method, device and equipment for processing multiple rounds of question-answering sentences and storage medium | |
| CN114328903A (en) | Text clustering-based customer service log backflow method and device | |
| Kumar et al. | An Algorithm for Automatic Text Annotation for Named Entity Recognition using spaCy Framework |
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 |