CN105843897A - Vertical domain-oriented intelligent question and answer system - Google Patents
Vertical domain-oriented intelligent question and answer system Download PDFInfo
- Publication number
- CN105843897A CN105843897A CN201610167602.5A CN201610167602A CN105843897A CN 105843897 A CN105843897 A CN 105843897A CN 201610167602 A CN201610167602 A CN 201610167602A CN 105843897 A CN105843897 A CN 105843897A
- Authority
- CN
- China
- Prior art keywords
- word
- module
- vocabulary
- sentence
- similarity
- 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
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
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Mathematical Physics (AREA)
- Human Computer Interaction (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Machine Translation (AREA)
Abstract
The invention discloses a vertical domain-oriented intelligent question and answer system. The system comprises a question asking module (1), a preprocessing module (2), a word segmentation and vocabulary standardization module (3), a word purification module (4), a synonym expansion module (5), a vocabulary expansion or deletion module (6), a sentence similarity calculation module (7) and an answer output module (8). The system calculates the similarity of question sentences of a user through domain ontology construction and depends on a word segmentation technology, domain ontology construction and ontology similarity calculation. The system has the advantages that a question asking intention of the user can be understood more accurately by applying a domain ontology technology through a sentence similarity algorithm, the sentence similarity can be calculated, and the accuracy of the question and answer system can be improved.
Description
Technical field
The present invention relates to a kind of Intelligent Answer System towards vertical field, the semantic analysis accuracy rate in vertical field is had
Significant and effect.
Background technology
Divide according to the technology that realizes of question answering system, including: question answering system based on frequently asked questions (FAQ), based on letter
The question answering system of breath retrieval, question answering system based on Question Classification and based on resource description framework (Resource
Description Framework) question answering system of RDF query.
Question answering system based on frequently asked questions, builds FAQs (FAQ) question and answer pair, it is achieved above depend on user's question sentence
With the Similarity Measure of question sentence in FAQ.In the development process of FAQ question answering system, need to identify the intention of user's question sentence, to two
Individual sentence carries out Similarity Measure, to return Query Result.The correlation technique flow process of existing FAQ question answering system is: to sentence
After carrying out participle, removing the pretreatment work such as stop words, word standardization, set up inverted index table, with VSM or TF-IDF algorithm
Calculate the similarity of the word array of two sentences.
Question answering system based on information retrieval, the information source of this system is typically the document on network, is returned
Answer is directly extracted from document.
Question answering system based on customer problem classification, generally builds corresponding template to each class problem and processes, increase
The strong understanding to problem, improves the accuracy rate of system.
Based on RDF(Resource Description Framework resource description framework, one is used for describing Web money
The markup language in source) core of question answering system inquired about is standard query language natural language question sentence being converted into RDF, generally
It is W3C given query language SPARQL, class, example or the attribute that the word in natural language question sentence is mapped as in body.
But prior art is when calculating Words similarity, there is employing similarity calculating method based on " knowing net ", but
Vertical field for specialty lacks enough semantic analysis.And prior art is when calculating sentence similarity, does not considers field
The weight of vocabulary, the vocabulary for the vertical field of specialty lacks enough semantic analysis.
The technical term explanation that the present invention relates to:
Domain body: domain body gives basic terminology and the relation constituting association area vocabulary, and combines these arts
Language and relation define the rule of these vocabulary extensions.
Participle technique: participle is exactly to be gone out by the words recognition of sentence and carry out part-of-speech tagging.
Know net: " knowing net " (HowNet) is the semantic knowledge dictionary that a comparison is detailed.With Chinese and english word institute's generation
The concept of table is description object, to disclose between concept and concept and relation between attribute that concept is had is in substantially
The commonsense knowledge base held.
Inverted index table: word is set up a table, and records the position of problem corresponding to word.Owing to not being by record
Determine property value, but determined the position of record by property value, thus referred to as inverted index (inverted index).
VSM: vector space model (Vector Space Model) is the process of content of text is reduced to vector space
In vector operation, the similarity of two vector operations is as the semantic similarity of two sentences.
TF-IDF: term frequency-inverse document frequency method (term frequency inverse document frequency),
On the basis of VSM algorithm, determine the weight of word according to the frequency of word, calculate the similarity of two sentences.
Summary of the invention
The present invention organically combines realization based on FAQ with based on RDF query technology, proposes a kind of new question answering system and process
Flow process, to strengthen Intelligent Answer System semantic analysis ability, improves the accuracy rate of intelligence automatically request-answering system.
The technical scheme is that the present invention calculates the similarity of user's question sentence by building domain body, depend on
Participle technique, the structure of domain body, body similarity calculate.
The invention have the advantage that, by this sentence similarity algorithm, application ontology understands use more accurately
Family is putd question to and is intended to, and calculates sentence similarity, improves the accuracy rate of question answering system.
Accompanying drawing explanation
Fig. 1 is that present system constitutes block diagram;
Fig. 2 is groundwork program flow diagram of the present invention;
Fig. 3 is the schematic diagram of the taxonomic structure embodiment of body of the present invention;
Fig. 4 is the structural representation of one concrete Noumenon property of the present invention;
Fig. 5 is the flow chart of one embodiment of working procedure of the present invention;
Fig. 6 is body baby's character classification by age structural representation of the present invention.
Detailed description of the invention
See Fig. 1, a kind of Intelligent Answer System towards vertical field of the present invention, it is based primarily upon computer system, including
Consisting of part:
(1) module 1 is putd question to: for inputting (proposition) problem to system.Input through keyboard, phonetic entry, hand-written (plate) can be used
Input, uses image collecting device input.
(2) pretreatment module 2: include vertical domain body (data base), for by the class in body, attribute, Instance Name
Claim to add in dictionary for word segmentation, and mark corresponding part of speech.
(3) participle and lexical normalisation module 3: for question sentence being carried out participle, and carrying out word standardization, marking each
Key words sorting in the part of speech of word and body.
(4) purify word module 4: for stop words is removed in the set after participle, remove the modal particle without practical significance,
Greeting word.
(5) synonym expansion module 5: for arranging the relevant Chinese thesaurus in vertical field, the meaning of a word is extended.
(6) ontology expansion module 6: for the lexical set after participle is judged, if the vocabulary in body, to word
Relation between remittance is analyzed, and is extended or deletes, and arranges this vocabulary weight in sentence;If not the word in body
Converge, calculate according to the similarity of common words.
(7) sentence similarity computing module: the weight in sentence of the vocabulary described in combination, calculates candidate in FAQ storehouse and asks
Topic and the sentence similarity of question sentence.
(8) output module is replied: for exporting the answer of problem.
Seeing Fig. 2, the groundwork flow process of the present invention includes:
(1) pretreatment: build vertical domain body, adds to the class in body, attribute, instance name in dictionary for word segmentation, and
Mark corresponding part of speech.
(2) question sentence carried out participle and carries out word standardization, marking the part of speech of each word, and the contingency table in body
Note.
(3) stop words is removed in the set after participle, remove the modal particle without practical significance, greeting word.
(4) arrange the relevant Chinese thesaurus in vertical field, the meaning of a word is extended.
(5) lexical set after participle is judged, if the vocabulary in body, the relation between vocabulary is carried out point
Analysis, is extended or deletes, and arranging this vocabulary weight in sentence;If not the vocabulary in body, according to common words
Similarity calculate.
(6) combine vocabulary weight in sentence, calculate candidate's problem and the sentence similarity of question sentence in FAQ storehouse.
(7) output problem answers: sort from high to low according to similarity, finally choose the highest problem of similarity as answering
Case.
Below in conjunction with Fig. 3-Fig. 6, system and the workflow of the present invention are described further.
1. build about vertical domain ontology repository:
The knowledge in vertical field is classified, the relation between concept of analysis and attribute thereof, it is achieved the expression of domain knowledge.
Class in domain body, example, attribute: class and example and class in object-oriented and to as if similar, attribute is retouched
State the relation between class or example.
In Fig. 4, " place ", as a class, has " Suzhou " example as it, has the reality of a Hui Shi golden clothes series
Example " Wyeth Hui Shi _ Promil Gold milk powder 2 sections 400g ", its place of production is Suzhou." place of production " connects two realities as attribute
Example.
2. the calculating of Words similarity in body:
Vocabulary is corresponding to class, example or the attribute in body.All concepts form directed graph, definition parent and direct subclass away from
From for 1, class is 1 with the distance of the example, and attribute defines the distance of territory and codomain and is respectively 1 with it, and the distance of vocabulary W1, W2 depends on
Add up according to above-mentioned definition.W0 is the nearest public father node of W1 and W2.The then semantic similarity employing formula of two vocabulary:
+
Such as Fig. 3: with " Thing " as root node, the degree of depth is 0, the degree of depth of " Wyeth Hui Shi _ Promil Gold milk powder 2 sections 400g "
It is 5, and the degree of depth of " Wyeth Hui Shi _ 3 sections of 400g of golden clothes Progress milk powder " is 5, their nearest public father node " golden clothes system
Row " the degree of depth be 4, then their similarity is+ =0.80。
Or:
α is an adjustable parameter, represents the value that two Lexical Similarities are its public father node of distance when 0.5.
Such as Fig. 3: set α=1.6, with " Thing " as root node, the degree of depth is 0, " Wyeth Hui Shi _ Promil Gold milk powder 2 sections
400g " and " Wyeth Hui Shi _ 3 sections of 400g of golden clothes Progress milk powder ", they away from nearest public father node " golden clothes series " away from
From being all 1, then their similarity is:
+ =0.62。
Finally, sort from high to low according to similarity, finally choose corresponding to first (similarity is the highest) problem
Answer is as the final result asked a question, and is exported by replying output module.
3. the determination of term weighing in question sentence:
The weights shared by word different in user's question sentence is different, and such as question sentence " may I ask colored king's diaper either with or without day basis
Dress?", the term weighing of " flower king's diaper " and " Japan is original-pack " be higher than " may I ask ", " having ", " not having ", " ".The most really
The method of determining is:
1) safeguard and disable vocabulary, will " " " " " " etc. get rid of without semantic word, be not counted in sentence similarity and calculate.
2) question sentence occurring, the situation that subclass is adjacent with its parent deletes parent.Chinese there will be the situation of semantic repetition,
Such as " Hui Shi milk powder " in Fig. 3, Hui Shi is the subclass of milk powder, and the information of subclass covers the information of parent, and the information that subclass is carried
Concrete in further detail, in this case, we only need to consider the information entrained by subclass.
3) analyze the dependence between word, if W1, W2 are modified relationship, and be the relation of Subject-Verb in the body
Then its object is added in vocabulary.
Such as Fig. 4: the SVO tlv triple in shown body, " Wyeth Hui Shi _ Promil Gold milk powder 2 sections 400g-product
Ground-Suzhou ", " Wyeth Hui Shi _ Promil Gold milk powder 2 sections 400g " is subject, and " place of production " is predicate, and " Suzhou " serves as object
Role.
Example 1: question sentence: " where the place of production of Wyeth Hui Shi _ 2 sections of 400g of Promil Gold milk powder is?", Chinese " A's
B ", in the case of A modifies B, " Suzhou " is added in vocabulary.
4) concept in domain body is the vocabulary higher with system degree of association, and along with the increase of the concept degree of depth, concept
The information carried is the most detailed, and therefore the term weight in domain knowledge is higher than normal words, and deep with vocabulary of the weight of vocabulary
Spend and increase.
Weightw0 = 1 +α
Wherein α is an adjustable parameter, the weight of regulation concept, and the value arranging α herein is 1, represents general in domain body
The weight read is between 1 and 2.
In Fig. 3, with " Thing " as root node, the degree of depth is 0, and the degree of depth of " Hui Shi " is 3, it may be assumed that=3,=5, the weight of word " Hui Shi " is WeightHui Shi = 1 + = 1.6。
5) calculate according to the number of words length of vocabulary:
Example 2: for the relation of milk powder Yu baby's age, the effect to semantic analysis of the ontology information of structure.
Such as Fig. 6, according to hop count and the age of applicable baby of milk powder, building corresponding domain body, system can identify " four
Month " be " 0-6 month " in the range of, thus find in model answer the problem containing " 0-6 month ", replying showing on output module
Show that content is as follows:
Such as Fig. 5, user inputs question sentence, and " why rice flour has rancid taste?", word is marked while participle by system
Standardization, is standardized as " abnormal flavour " by " rancid taste ";Remove stop words, " " " will be had " to remove;Look into data base's inverted index table
In containing the question sentence of [reason, rice flour, abnormal flavour], and problem is sorted according to the quantity containing key word, takes front 15 question sentences and make
For candidate's problem;Use VSM algorithm, calculate successively the participle of these 15 candidate's problems go stop words result with [reason, rice flour,
Abnormal flavour] similarity, sequence;Similarity is chosen the answer of the highest problem and is returned.
Claims (3)
1. the Intelligent Answer System towards vertical field, it is characterised in that include consisting of part:
(1) module is putd question to: for inputting problem to system;
(2) pretreatment module: include vertical domain body, for adding the class in body, attribute, instance name to participle word
In allusion quotation, and mark corresponding part of speech;
(3) participle and lexical normalisation module: for question sentence is carried out participle, marks the part of speech of each word, and dividing in body
Class labelling;
(4) word module is purified: for stop words is removed in the set after participle, remove the modal particle without practical significance, greeting
Word;
(5) synonym expansion module: for arranging the relevant Chinese thesaurus in vertical field, the meaning of a word is extended;
(6) ontology expansion module: for the lexical set after participle is judged, if the vocabulary in body, between vocabulary
Relation be analyzed, be extended or delete, and this vocabulary weight in sentence is set;If not the vocabulary in body,
Calculate according to the similarity of common words;
(7) sentence similarity computing module: the weight in sentence of the vocabulary described in combination, calculate in FAQ storehouse candidate's problem with
The sentence similarity of question sentence;
(8) output module is replied: for exporting the answer of problem.
Intelligent Answer System towards vertical field the most according to claim 1, it is characterised in that described enquirement module
Use keyboard, voice, hand-written or image collecting device input;Described answer output module uses display, speaker or beats
Print machine.
Intelligent Answer System towards vertical field the most according to claim 1, it is characterised in that the workflow of this system
Journey includes:
(1) pretreatment: build vertical domain body, adds to the class in body, attribute, instance name in dictionary for word segmentation, and
Mark corresponding part of speech;
(2) question sentence carried out participle and carries out word standardization, marking the part of speech of each word, and the key words sorting in body;
(3) stop words is removed in the set after participle, remove the modal particle without practical significance, greeting word;
(4) arrange the relevant Chinese thesaurus in vertical field, the meaning of a word is extended;
(5) lexical set after participle is judged, if the vocabulary in body, the relation between vocabulary is analyzed, enters
Row extension or deletion, and this vocabulary weight in sentence is set;If not the vocabulary in body, similar according to common words
Degree calculates;
(6) combine vocabulary weight in sentence, calculate candidate's problem and the sentence similarity of question sentence in FAQ storehouse;
(7) output problem answers: sort from high to low according to similarity, finally choose answering corresponding to the problem that similarity is the highest
Case is as the answer of problem.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610167602.5A CN105843897B (en) | 2016-03-23 | 2016-03-23 | A kind of intelligent Answer System towards vertical field |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610167602.5A CN105843897B (en) | 2016-03-23 | 2016-03-23 | A kind of intelligent Answer System towards vertical field |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105843897A true CN105843897A (en) | 2016-08-10 |
CN105843897B CN105843897B (en) | 2019-03-26 |
Family
ID=56584443
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610167602.5A Active CN105843897B (en) | 2016-03-23 | 2016-03-23 | A kind of intelligent Answer System towards vertical field |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105843897B (en) |
Cited By (23)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106503101A (en) * | 2016-10-14 | 2017-03-15 | 五邑大学 | Electric business customer service automatically request-answering system sentence keyword extracting method |
CN106874441A (en) * | 2017-02-07 | 2017-06-20 | 腾讯科技(上海)有限公司 | Intelligent answer method and apparatus |
CN107657575A (en) * | 2017-09-30 | 2018-02-02 | 四川智美高科科技有限公司 | A kind of government affairs Intelligent Service terminal device and application method based on artificial intelligence |
CN107679231A (en) * | 2017-10-24 | 2018-02-09 | 济南浪潮高新科技投资发展有限公司 | A kind of vertical field and the implementation method of Opening field mixed type intelligent Answer System |
CN107679039A (en) * | 2017-10-17 | 2018-02-09 | 北京百度网讯科技有限公司 | The method and apparatus being intended to for determining sentence |
CN107688608A (en) * | 2017-07-28 | 2018-02-13 | 合肥美的智能科技有限公司 | Intelligent sound answering method, device, computer equipment and readable storage medium storing program for executing |
CN107688667A (en) * | 2017-09-30 | 2018-02-13 | 平安科技(深圳)有限公司 | Intelligent robot client service method, electronic installation and computer-readable recording medium |
CN107704559A (en) * | 2017-09-29 | 2018-02-16 | 软通动力信息技术(集团)有限公司 | A kind of semantic understanding method and device |
CN108121800A (en) * | 2017-12-21 | 2018-06-05 | 北京百度网讯科技有限公司 | Information generating method and device based on artificial intelligence |
CN108256009A (en) * | 2018-01-03 | 2018-07-06 | 国网江苏省电力有限公司电力科学研究院 | A kind of method for improving electric intelligent response robot and answering accuracy rate |
CN108509416A (en) * | 2018-03-20 | 2018-09-07 | 京东方科技集团股份有限公司 | Sentence realizes other method and device, equipment and storage medium |
CN108628882A (en) * | 2017-03-20 | 2018-10-09 | 北京京东尚科信息技术有限公司 | Method and system for prejudging problem |
WO2019080910A1 (en) * | 2017-10-25 | 2019-05-02 | 中兴通讯股份有限公司 | Information processing system and method thereof for implementing information processing |
CN109710732A (en) * | 2018-11-19 | 2019-05-03 | 东软集团股份有限公司 | Information query method, device, storage medium and electronic equipment |
CN110532566A (en) * | 2019-09-03 | 2019-12-03 | 山东浪潮通软信息科技有限公司 | A kind of implementation method that vertical field Question sentence parsing calculates |
CN110866089A (en) * | 2019-11-14 | 2020-03-06 | 国家电网有限公司 | Robot knowledge base construction system and method based on synonymous multi-language environment analysis |
CN111324721A (en) * | 2020-03-16 | 2020-06-23 | 云南电网有限责任公司信息中心 | Method for constructing intelligent question-answering knowledge base |
CN112256864A (en) * | 2020-09-23 | 2021-01-22 | 北京捷通华声科技股份有限公司 | Multi-intention recognition method and device, electronic equipment and readable storage medium |
CN112307738A (en) * | 2020-11-11 | 2021-02-02 | 北京沃东天骏信息技术有限公司 | Method and device for processing text |
CN112613611A (en) * | 2020-12-29 | 2021-04-06 | 上海永骁智能技术有限公司 | Tax knowledge base system based on knowledge graph |
CN112668324A (en) * | 2020-12-04 | 2021-04-16 | 北京达佳互联信息技术有限公司 | Corpus data processing method and device, electronic equipment and storage medium |
CN113360596A (en) * | 2021-06-08 | 2021-09-07 | 南京大学 | Crowd-sourcing assistant implementation method based on tf-idf and inverted index |
CN116975206A (en) * | 2023-09-25 | 2023-10-31 | 华云天下(南京)科技有限公司 | Vertical field training method and device based on AIGC large model and electronic equipment |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101373532A (en) * | 2008-07-10 | 2009-02-25 | 昆明理工大学 | FAQ Chinese request-answering system implementing method in tourism field |
-
2016
- 2016-03-23 CN CN201610167602.5A patent/CN105843897B/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101373532A (en) * | 2008-07-10 | 2009-02-25 | 昆明理工大学 | FAQ Chinese request-answering system implementing method in tourism field |
Non-Patent Citations (2)
Title |
---|
周永梅等: "自动问答系统中的句子相似度算法的研究", 《计算机技术与发展》 * |
袁青波: "基于本体的语义相似度计算在营养知识问答系统中的应用", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (34)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106503101A (en) * | 2016-10-14 | 2017-03-15 | 五邑大学 | Electric business customer service automatically request-answering system sentence keyword extracting method |
CN106874441A (en) * | 2017-02-07 | 2017-06-20 | 腾讯科技(上海)有限公司 | Intelligent answer method and apparatus |
CN106874441B (en) * | 2017-02-07 | 2024-03-05 | 腾讯科技(上海)有限公司 | Intelligent question-answering method and device |
CN108628882A (en) * | 2017-03-20 | 2018-10-09 | 北京京东尚科信息技术有限公司 | Method and system for prejudging problem |
CN107688608A (en) * | 2017-07-28 | 2018-02-13 | 合肥美的智能科技有限公司 | Intelligent sound answering method, device, computer equipment and readable storage medium storing program for executing |
CN107704559A (en) * | 2017-09-29 | 2018-02-16 | 软通动力信息技术(集团)有限公司 | A kind of semantic understanding method and device |
CN107704559B (en) * | 2017-09-29 | 2020-06-05 | 北京软通智城科技有限公司 | Semantic understanding method and device |
CN107688667A (en) * | 2017-09-30 | 2018-02-13 | 平安科技(深圳)有限公司 | Intelligent robot client service method, electronic installation and computer-readable recording medium |
CN107657575A (en) * | 2017-09-30 | 2018-02-02 | 四川智美高科科技有限公司 | A kind of government affairs Intelligent Service terminal device and application method based on artificial intelligence |
CN107679039A (en) * | 2017-10-17 | 2018-02-09 | 北京百度网讯科技有限公司 | The method and apparatus being intended to for determining sentence |
CN107679039B (en) * | 2017-10-17 | 2020-12-29 | 北京百度网讯科技有限公司 | Method and device for determining statement intention |
CN107679231A (en) * | 2017-10-24 | 2018-02-09 | 济南浪潮高新科技投资发展有限公司 | A kind of vertical field and the implementation method of Opening field mixed type intelligent Answer System |
WO2019080910A1 (en) * | 2017-10-25 | 2019-05-02 | 中兴通讯股份有限公司 | Information processing system and method thereof for implementing information processing |
CN108121800A (en) * | 2017-12-21 | 2018-06-05 | 北京百度网讯科技有限公司 | Information generating method and device based on artificial intelligence |
CN108121800B (en) * | 2017-12-21 | 2021-12-21 | 北京百度网讯科技有限公司 | Information generation method and device based on artificial intelligence |
CN108256009A (en) * | 2018-01-03 | 2018-07-06 | 国网江苏省电力有限公司电力科学研究院 | A kind of method for improving electric intelligent response robot and answering accuracy rate |
CN108256009B (en) * | 2018-01-03 | 2022-02-15 | 国网江苏省电力有限公司电力科学研究院 | Method for improving answer accuracy of electric intelligent answer robot |
CN108509416A (en) * | 2018-03-20 | 2018-09-07 | 京东方科技集团股份有限公司 | Sentence realizes other method and device, equipment and storage medium |
CN109710732B (en) * | 2018-11-19 | 2021-03-05 | 东软集团股份有限公司 | Information query method, device, storage medium and electronic equipment |
CN109710732A (en) * | 2018-11-19 | 2019-05-03 | 东软集团股份有限公司 | Information query method, device, storage medium and electronic equipment |
CN110532566A (en) * | 2019-09-03 | 2019-12-03 | 山东浪潮通软信息科技有限公司 | A kind of implementation method that vertical field Question sentence parsing calculates |
CN110532566B (en) * | 2019-09-03 | 2023-05-02 | 浪潮通用软件有限公司 | Method for realizing similarity calculation of questions in vertical field |
CN110866089A (en) * | 2019-11-14 | 2020-03-06 | 国家电网有限公司 | Robot knowledge base construction system and method based on synonymous multi-language environment analysis |
CN110866089B (en) * | 2019-11-14 | 2023-04-28 | 国家电网有限公司 | Robot knowledge base construction system and method based on synonymous multi-context analysis |
CN111324721A (en) * | 2020-03-16 | 2020-06-23 | 云南电网有限责任公司信息中心 | Method for constructing intelligent question-answering knowledge base |
CN112256864A (en) * | 2020-09-23 | 2021-01-22 | 北京捷通华声科技股份有限公司 | Multi-intention recognition method and device, electronic equipment and readable storage medium |
CN112256864B (en) * | 2020-09-23 | 2024-05-14 | 北京捷通华声科技股份有限公司 | Multi-intention recognition method, device, electronic equipment and readable storage medium |
CN112307738A (en) * | 2020-11-11 | 2021-02-02 | 北京沃东天骏信息技术有限公司 | Method and device for processing text |
CN112668324A (en) * | 2020-12-04 | 2021-04-16 | 北京达佳互联信息技术有限公司 | Corpus data processing method and device, electronic equipment and storage medium |
CN112668324B (en) * | 2020-12-04 | 2023-12-08 | 北京达佳互联信息技术有限公司 | Corpus data processing method and device, electronic equipment and storage medium |
CN112613611A (en) * | 2020-12-29 | 2021-04-06 | 上海永骁智能技术有限公司 | Tax knowledge base system based on knowledge graph |
CN113360596A (en) * | 2021-06-08 | 2021-09-07 | 南京大学 | Crowd-sourcing assistant implementation method based on tf-idf and inverted index |
CN116975206B (en) * | 2023-09-25 | 2023-12-08 | 华云天下(南京)科技有限公司 | Vertical field training method and device based on AIGC large model and electronic equipment |
CN116975206A (en) * | 2023-09-25 | 2023-10-31 | 华云天下(南京)科技有限公司 | Vertical field training method and device based on AIGC large model and electronic equipment |
Also Published As
Publication number | Publication date |
---|---|
CN105843897B (en) | 2019-03-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105843897B (en) | A kind of intelligent Answer System towards vertical field | |
CN103678564B (en) | Internet product research system based on data mining | |
US9613024B1 (en) | System and methods for creating datasets representing words and objects | |
CN107239439A (en) | Public sentiment sentiment classification method based on word2vec | |
CN102831184B (en) | According to the method and system text description of social event being predicted to social affection | |
US9880998B1 (en) | Producing datasets for representing terms and objects based on automated learning from text contents | |
CN103207913B (en) | The acquisition methods of commercial fine granularity semantic relation and system | |
CN106776562A (en) | A kind of keyword extracting method and extraction system | |
CN106156272A (en) | A kind of information retrieval method based on multi-source semantic analysis | |
CN110674252A (en) | High-precision semantic search system for judicial domain | |
CN107992542A (en) | A kind of similar article based on topic model recommends method | |
CN110362678A (en) | A kind of method and apparatus automatically extracting Chinese text keyword | |
CN105843796A (en) | Microblog emotional tendency analysis method and device | |
CN109960756A (en) | Media event information inductive method | |
CN108388554A (en) | Text emotion identifying system based on collaborative filtering attention mechanism | |
CN109255012A (en) | A kind of machine reads the implementation method and device of understanding | |
CN108804595A (en) | A kind of short text representation method based on word2vec | |
Jayaram et al. | A review: Information extraction techniques from research papers | |
Haque et al. | Opinion mining from bangla and phonetic bangla reviews using vectorization methods | |
Sadr et al. | Unified topic-based semantic models: A study in computing the semantic relatedness of geographic terms | |
CN106055633A (en) | Chinese microblog subjective and objective sentence classification method | |
JPWO2014002774A1 (en) | Synonym extraction system, method and recording medium | |
Xu et al. | A classification of questions using SVM and semantic similarity analysis | |
Ji et al. | Research on semantic similarity calculation methods in Chinese financial intelligent customer service | |
CN113298365B (en) | Cultural additional value assessment method based on LSTM |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CP03 | Change of name, title or address | ||
CP03 | Change of name, title or address |
Address after: Floor 5, Haier cloud Valley innovation block a, No. 1, Haier Road, Laoshan District, Qingdao, Shandong Province Patentee after: QINGDAO PENGHAI SOFTWARE Co.,Ltd. Address before: Laoshan District Green Road 266071 Shandong city of Qingdao province No. 19 22 11 floor Patentee before: HAIERSOFT Corp. |