CN116414961A - Question-answering method and system based on military domain knowledge graph - Google Patents
Question-answering method and system based on military domain knowledge graph Download PDFInfo
- Publication number
- CN116414961A CN116414961A CN202310281351.3A CN202310281351A CN116414961A CN 116414961 A CN116414961 A CN 116414961A CN 202310281351 A CN202310281351 A CN 202310281351A CN 116414961 A CN116414961 A CN 116414961A
- Authority
- CN
- China
- Prior art keywords
- question
- data
- user
- intention
- text
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 66
- 238000012545 processing Methods 0.000 claims abstract description 23
- 238000004590 computer program Methods 0.000 claims description 17
- 230000008569 process Effects 0.000 claims description 17
- 238000003860 storage Methods 0.000 claims description 16
- 230000000007 visual effect Effects 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 230000001976 improved effect Effects 0.000 abstract description 5
- 238000000605 extraction Methods 0.000 description 22
- 238000007726 management method Methods 0.000 description 21
- 238000010586 diagram Methods 0.000 description 16
- 230000006870 function Effects 0.000 description 14
- 230000000875 corresponding effect Effects 0.000 description 13
- 230000004927 fusion Effects 0.000 description 13
- 230000003993 interaction Effects 0.000 description 10
- 238000012549 training Methods 0.000 description 9
- 238000012423 maintenance Methods 0.000 description 8
- 238000012795 verification Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 238000001914 filtration Methods 0.000 description 6
- 238000006243 chemical reaction Methods 0.000 description 5
- 238000010276 construction Methods 0.000 description 5
- 238000012217 deletion Methods 0.000 description 5
- 230000037430 deletion Effects 0.000 description 5
- 238000009826 distribution Methods 0.000 description 5
- 230000010354 integration Effects 0.000 description 5
- 238000013508 migration Methods 0.000 description 5
- 230000005012 migration Effects 0.000 description 5
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000001360 synchronised effect Effects 0.000 description 4
- 238000012546 transfer Methods 0.000 description 4
- 238000012800 visualization Methods 0.000 description 4
- 238000001514 detection method Methods 0.000 description 3
- 239000012634 fragment Substances 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000007792 addition Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 101150026173 ARG2 gene Proteins 0.000 description 1
- 101100005166 Hypocrea virens cpa1 gene Proteins 0.000 description 1
- 235000010627 Phaseolus vulgaris Nutrition 0.000 description 1
- 244000046052 Phaseolus vulgaris Species 0.000 description 1
- 101100379633 Xenopus laevis arg2-a gene Proteins 0.000 description 1
- 101100379634 Xenopus laevis arg2-b gene Proteins 0.000 description 1
- 230000003213 activating effect Effects 0.000 description 1
- 101150088826 arg1 gene Proteins 0.000 description 1
- 238000012098 association analyses Methods 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 239000000969 carrier Substances 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 238000013524 data verification Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000011049 filling Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 230000009191 jumping Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000011068 loading method Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000005055 memory storage Effects 0.000 description 1
- 230000035515 penetration Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000012913 prioritisation Methods 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
- 230000008093 supporting effect Effects 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—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
- 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
- G06F16/3343—Query execution using phonetics
-
- 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
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
- G06F40/295—Named entity recognition
-
- 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
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
- G06N5/022—Knowledge engineering; Knowledge acquisition
- G06N5/025—Extracting rules from data
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Databases & Information Systems (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Acoustics & Sound (AREA)
- Life Sciences & Earth Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application relates to a question-answering method and system based on a military domain knowledge graph, wherein the question-answering method based on the military domain knowledge graph comprises the following steps: acquiring a user voice question, and performing voice recognition on the question to generate a corresponding text, or acquiring a user text question to obtain a corresponding text; determining the semantics of the text to obtain the intention of a user, and determining answer content based on a pre-constructed military domain knowledge base according to the intention; through the application, the efficiency and accuracy of people for acquiring and processing military field information are improved.
Description
Technical Field
The application relates to the technical field of data processing, in particular to a question-answering method and system based on a military domain knowledge graph.
Background
The military intelligent penetration is carried out to each link of a military system, the traditional information processing mode can not meet the understanding requirement of military personnel on military knowledge, the military information which is open in source on the network comprises various forms such as videos, voices, characters, pictures, tables and the like, the information is finely divided and segmented, valuable information and interrelationships thereof need to be found from massive data, the time for people to acquire the useful information is greatly prolonged, and the information acquired in different channels is not uniform. Meanwhile, the military field lacks of knowledge graph products, the application scene of the existing knowledge graph products lacks of deep research, the knowledge graph products cannot be deeply applied to intelligent question-answering service in the military field, and business personnel lack of favorable tools for acquiring information and processing data.
Aiming at the problem of low efficiency and accuracy of people in acquiring and processing information in the military field in the related art, no effective solution has been proposed yet.
Disclosure of Invention
The embodiment of the application provides a question-answering method and system based on a military field knowledge graph, which at least solve the problem that in the related art, the efficiency and accuracy of people for acquiring and processing the military field information are low.
In a first aspect, an embodiment of the present application provides a question-answering method based on a knowledge graph in the military field, where the method includes:
acquiring a user voice question, and performing voice recognition on the question to generate a corresponding text, or acquiring a user text question to obtain a corresponding text;
determining the semantics of the text, obtaining the intention of a user, and determining answer content based on a pre-constructed military domain knowledge base according to the intention.
In some of these embodiments, determining the text semantics includes:
extracting semantic elements from the text, and judging the intention of a question based on a preset rule or model according to the extracted semantic elements;
and according to the intention discrimination result, converting the question semantic elements into system query sentences so as to query the answer content in the military domain knowledge base.
In some embodiments, in the event that the user intent cannot be determined, before the converting the question semantic elements to the system query statement, the method includes:
and distributing the intention to a plurality of task scenes to develop a multi-round dialogue, wherein in one multi-round process, if the user expresses the intention of another task scene, the user enters into another task scene, and the plurality of task scenes share word slot information.
In some of these embodiments, the process of determining answer content includes:
determining a query result through a multi-engine question-answer strategy, sequencing, screening and merging the query result, determining answer content of a user question, and displaying the answer content in a visual form;
wherein, the multi-engine question-answering strategy comprises: and based on knowledge graph questions and answers, reading and understanding the questions and answers, and the questions and answers are versus library questions and answers, performing retrieval in the military domain knowledge base to obtain the query result.
In some embodiments, the construction process of the military domain knowledge base includes:
the method comprises the steps of leading and integrating data, and integrating the data through a data processing engine, wherein the data comprises business data, open source data and third party data;
And extracting the target entity, the entity attribute and the naming relation among the entities from the data, fusing and disambiguating the extracted knowledge, and storing the knowledge in the military domain knowledge base.
In a second aspect, an embodiment of the present application provides a question-answering system based on a military domain knowledge graph, where the system includes:
the acquisition module is used for acquiring a user voice question and carrying out voice recognition on the question to generate a corresponding text, or acquiring a user text question to obtain the corresponding text;
and the determining module is used for determining the semantics of the text, obtaining the intention of the user, and determining the answer content based on a pre-constructed military domain knowledge base according to the intention.
In some of these embodiments, in the determining module, the determining the text semantics includes:
extracting semantic elements from the text, and judging the intention of a question based on a preset rule or model according to the extracted semantic elements; and according to the intention discrimination result, converting the question semantic elements into system query sentences so as to query the answer content in the military domain knowledge base.
In some embodiments, the determining module is further configured to, in a case where the user intent cannot be determined, prior to converting the question semantic element into the system query statement:
And distributing the intention to a plurality of task scenes to develop a multi-round dialogue, wherein in one multi-round process, if the user expresses the intention of another task scene, the user enters into another task scene, and the plurality of task scenes share word slot information.
In a third aspect, embodiments of the present application provide an electronic device comprising a memory, and a processor, the memory having stored therein a computer program, the processor being configured to run the computer program to perform the military domain knowledge graph based question-answering method.
In a fourth aspect, embodiments of the present application provide a storage medium having a computer program stored therein, where the computer program is configured to execute the military domain knowledge graph-based question-answering method at runtime.
Compared with the related art, the question-answering method based on the military domain knowledge graph provided by the embodiment of the application has the advantages that through obtaining user input, analyzing user intention, determining answer content based on the pre-constructed military domain knowledge base, and improving efficiency and accuracy of people for obtaining and processing military domain information;
in addition, according to the service requirements and the data actual conditions, the embodiment of the application builds a professional knowledge base in the field, realizes the deep integration of the data, fuses the isolated fragments of the data into a panoramic jigsaw, and effectively improves the data resource utilization capacity;
The method comprises the steps of accurately judging the intention of a field problem by means of machine learning, semantic understanding and the like, establishing a multi-engine question-answer strategy, rapidly inquiring and searching a knowledge base by means of knowledge map question-answer, reading and understanding question-answer, question-answer and library question-answer and the like, sorting, screening and merging inquiry results, and displaying the question-answer results in various visual forms; when the intention of the problem is fuzzy, the focus requirement of the user is guided through multiple rounds of conversations, related problems and results are recommended, and the experience degree of the user and the accuracy of information acquisition are greatly improved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute an undue limitation to the application. In the drawings:
FIG. 1 is a schematic view of an application environment of a question-answering method based on a military domain knowledge graph according to an embodiment of the present application;
FIG. 2 is a schematic diagram of a questioning and answering workflow based on military domain knowledge base according to a first embodiment of the present application;
FIG. 3 is a flow chart of a method of question-answering based on a military domain knowledge-graph according to a second embodiment of the present application;
FIG. 4 is a schematic diagram of the operation of a system for intelligent question-answering based on knowledge-graph of military domain according to a third embodiment of the present application;
FIG. 5 is a platform architecture diagram of a military domain knowledge base in accordance with a fourth embodiment of the present application;
FIG. 6 is a general architecture diagram of a military field question-answering system according to a fifth embodiment of the present application;
FIG. 7 is a flowchart of a construction process of a military domain knowledge base in accordance with a sixth embodiment of the present application;
FIG. 8 is a knowledge fusion architecture diagram of a military domain knowledge base in accordance with a seventh embodiment of the application;
fig. 9 is a schematic diagram of an internal structure of an electronic device according to an eighth embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described and illustrated below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application. All other embodiments, which can be made by one of ordinary skill in the art without undue burden on the person of ordinary skill in the art based on the embodiments provided herein, are intended to be within the scope of the present application.
It is apparent that the drawings in the following description are only some examples or embodiments of the present application, and it is possible for those of ordinary skill in the art to apply the present application to other similar situations according to these drawings without inventive effort. Moreover, it should be appreciated that while such a development effort might be complex and lengthy, it would nevertheless be a routine undertaking of design, fabrication, or manufacture for those of ordinary skill having the benefit of this disclosure, and thus should not be construed as having the benefit of this disclosure.
Reference in the specification to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is to be expressly and implicitly understood by those of ordinary skill in the art that the embodiments described herein can be combined with other embodiments without conflict.
Unless defined otherwise, technical or scientific terms used herein should be given the ordinary meaning as understood by one of ordinary skill in the art to which this application belongs. Reference to "a," "an," "the," and similar terms herein do not denote a limitation of quantity, but rather denote the singular or plural. The terms "comprising," "including," "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, article, or apparatus that comprises a list of steps or modules (elements) is not limited to only those steps or elements but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus. The terms "connected," "coupled," and the like in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The term "plurality" as used herein refers to two or more. "and/or" describes an association relationship of an association object, meaning that there may be three relationships, e.g., "a and/or B" may mean: a exists alone, A and B exist together, and B exists alone. The character "/" generally indicates that the context-dependent object is an "or" relationship. The terms "first," "second," "third," and the like, as used herein, are merely distinguishing between similar objects and not representing a particular ordering of objects.
The question-answering method based on the military domain knowledge graph provided by the application can be applied to an application environment shown in fig. 1, fig. 1 is a schematic diagram of the application environment of the question-answering method based on the military domain knowledge graph according to the embodiment of the application, and as shown in fig. 1, a terminal 102 and a server 104 communicate through a network. The server 104 obtains a user voice question through the terminal 102 and performs voice recognition on the question to generate a corresponding text, or the server 104 obtains the user text question through the terminal 102 to obtain the corresponding text; the server 104 determines the semantics of the text to obtain the intention of the user, and determines answer content based on a pre-constructed military domain knowledge base according to the intention; the terminal 102 presents the answer content to the user. The terminal 102 has a voice wake-up function, can naturally interact with a user in a wake-up state, provides various interaction modes such as voice dialogue, text input and the like, supports visual presentation of questions and answers, and the terminal 102 can be, but is not limited to, a variety of man-machine interaction terminals such as personal computers, notebook computers, smart phones, intelligent robots, intelligent voice assistants, tablet computers, portable wearable devices and the like, and the server 104 can be implemented by an independent server or a server cluster formed by a plurality of servers.
The embodiment provides a question-answering method based on a military domain knowledge graph, fig. 2 is a schematic diagram of a question-answering workflow based on the military domain knowledge graph according to the first embodiment of the application, and as shown in fig. 2, the question-answering workflow relates to links such as man-machine interaction, question sentence semantic understanding, multi-party collaborative question-answering and the like; fig. 3 is a flowchart of a question-answering method based on a military domain knowledge graph according to a second embodiment of the present application, as shown in fig. 3, the flowchart includes the steps of:
step S301, a user voice question is obtained, voice recognition is carried out on the question, and a corresponding text is generated, or a user text question is obtained, and the corresponding text is obtained;
for example, acquiring a user voice question, carrying out accent correction processing, accurately recognizing voice under the conditions of no large-scale grammar difference and noise to a certain extent, and realizing text conversion based on a Chinese voice intelligent recognition model, wherein the voice recognition model supports an online learning function, converts incremental data into training samples (manual supervision), and continuously optimizes model effects, so that the method is suitable for business application scenes, and realizes rapid and accurate conversion from voice to text;
optimizing the voice recognition result, optimizing the recognition result based on Chinese language rules and an optimization model, generating a question text of natural language description, displaying the question text on the terminal 102, and storing and recording original audio and the recognition result; the optimization model can be learned and updated online, so that incremental data are converted into training samples, and the model effect is continuously improved;
Step S302, determining the semantics of the text to obtain the intention of the user; namely, on the basis of voice recognition, the field problem is accurately judged, spoken language with specific context is converted into query sentences which can be recognized by a system, and the field problem intention is accurately judged through methods such as machine learning, semantic understanding and the like; for example, through natural language understanding related technology, judging what a user wants to accomplish or what the user wants to put forward, extracting semantic elements, judging the intention of the problem based on rules or models according to the extracted semantic elements, and converting question semantic elements into system query sentences according to the intention judging result;
the specific processing procedure for intention determination is as follows:
1) Filtering stop words and correcting professional words; filtering stop words such as o, what and the like in the problem presented by the visitor, and setting a professional vocabulary management page which comprises the steps of adding synonyms, editing the synonyms, deleting the synonyms and the like;
2) Extracting semantic elements; the entity identification comprises a preset entity and a custom entity, wherein the custom entity comprises an enumeration entity, a regular entity and an intention entity; extracting key information from a message sent by a user, wherein the key information corresponds to an entity and an entity value, the entity and the entity value are used for filling a slot in a task robot, the word slot can refer to the entity, the condition that the user speaks which fragments can be identified as the value of the word slot is specified, the identification mode comprises keyword matching, rule matching and the like, the user can inquire and modify the condition of all the entity and the entity value of the current project, and the entity identification algorithm is realized by adopting a flexible configuration method of three algorithms, namely a conditional random field-based algorithm, a BiLSTM+CRF-based algorithm and a BERT-based algorithm;
3) Judging intention; the user questioning is subjected to similarity matching with the similarity speaking methods of all intentions, so that the user intentions are identified, the intention triggering mode comprises keywords and the similarity speaking methods, and the keywords support two triggering rules of 'equal' and 'comprising': in the "equal" condition, it is active when the user's input and settings match exactly the keywords; in the 'containing' condition, when keywords are contained in the input of a user, the 'similar expression' is based on semantic similarity, the intention triggering condition of the similar expression is matched, an intention recognition algorithm is realized by adopting a search model, a knowledge base consisting of a large number of query-response sentence pairs is firstly constructed, sources comprise chat sentence pairs of social networks such as bar pasting, bean paste, microblog and the like or domain knowledge produced by a manual customer service system, the last reply in the dialogue is taken as a query, and the correlation between the query-query is calculated by an information search mode (inverted index+TFIDF/BM 25) to recall candidate responses;
it is worth to say that the platform presets some intents to help the user to realize a general scene, the preset intents comprise no intents and push-out intents, the user can also manually create the intents, and after the intention is created successfully, the management page displays all the intention lists, wherein the intention lists comprise intention names, intention triggering modes and intention use cases;
4) Generating a query instruction, wherein a common user can complete the query work of a complex database through natural language description to obtain a desired result, and the instruction generation adopts a Text2SQL algorithm which comprises BART (two-way autoregressive pre-training model) and GraPPA (grammar enhanced pre-training model for table semantic analysis); in addition, adopting an AdaBoost method to classify the question text;
step S303, determining answer content based on a pre-constructed military domain knowledge base according to the intention; optionally, a multi-engine question-answer strategy can be established, the knowledge base is rapidly searched and searched in a plurality of modes such as knowledge map question-answer, reading and understanding question-answer, question-answer and library question-answer, and the query results are sequenced, screened and combined and the question-answer results are displayed in a plurality of visual modes; in addition, when the intention of the problem is fuzzy, the focusing requirement of the user can be guided through multiple rounds of conversations, and related problems and results are recommended, so that the experience of the user and the accuracy of information acquisition are greatly improved;
for example, fig. 4 is a schematic diagram of the operation of the intelligent question-answering system based on the military domain knowledge graph according to the third embodiment of the present application, as shown in fig. 4, the central control is a transfer station of each downstream robot, and after the request comes, the central control calls the NLU module to perform intent analysis and Query preprocessing, then distributes the request to the downstream service question-answering robot, the graph robot and the MRC robot in parallel, asynchronously receives the return of the downstream service, and determines the final adoption result in real time based on the caching of the history session by the central control; the process specifically comprises the following steps: multi-engine collaborative question-answering strategy, knowledge reasoning and retrieval, question-answering result filtering and sorting, user demand feedback, new knowledge point learning, personalized question-answering and model training and the like:
1) A multi-engine collaborative question-answering strategy; the flexible skip of a plurality of robot works or a plurality of multi-round dialogue scenes is realized through intention distribution, and a robot system can flexibly skip to a corresponding intention triggering scene according to the intention of a user in a multi-round interaction scene; the central control platform is responsible for: i. the intention distribution is carried out, and the intention is analyzed and understood through an intention engine according to user input and then is forwarded to a corresponding service robot; when the intention of the user is not confirmed, the user is supported to be asked reversely; ii. In a multi-round state management, if a client expresses the intention of a multi-round task scene, entering another multi-round task scene, recording the current jump state by a central control platform, and after the jump-in multi-round task is completed, returning to the previous multi-round dialogue through the state record of the central control platform, and recovering all states before jumping out; iii, sharing word slots in multiple rounds, wherein when the task type jump of one robot or a plurality of robots is involved, the two corresponding task type scenes have common word slot information, the word slot information can be shared, and corresponding actions are carried out according to the customer expression after the word slot collection is completed; different key values can be redefined to realize information sharing related to a third party robot; when many robot scenes are needed, the robots are needed to manage the multiple robots and distribute the flow, and the central control dialogue supports: i. word slot sharing, which provides information for different robots and different multi-round processes to realize word slot information sharing; ii. Policy configuration, which can set allocation policies of different outgoing modules, including scores, priorities, thresholds, etc.; iii, managing the outgoing call modules, wherein the outgoing call modules of the robots are managed uniformly, and the management operation comprises adding, activating, editing and the like;
2) Knowledge reasoning and searching; knowledge graph-oriented reasoning mainly surrounds the reasoning expansion of the relation, namely unknown facts or relations are deduced based on the existing facts or relations in the graph, and features of three aspects of entities, relations and graph structures are generally emphasized and inspected; in general, knowledge graph reasoning can assist in reasoning out new facts, relationships, axioms and new rules; the main functions of reasoning are: the knowledge graph is subjected to complementation, quality check, link prediction, association relation reasoning, conflict detection and the like through rule mining;
3) Filtering and sequencing the question and answer results; dialog prioritization and threshold filtering; i. dialog priority refers to which one should be replied with priority when the robot matches multiple dialog types for the same sentence of user messages; the reply mode refers to whether replies of various dialogue types are automatically sent out, and the reply mode provides three modes: designating priority, full reply and no reply; ii. In the threshold setting, thresholds for question-answer dialog, atlas dialog, and MRC dialog may be set: when the confidence coefficient of the similarity of one knowledge point exceeds the set question-answer threshold, the knowledge point similarity is selected into the candidate replies of the robot; when the confidence level of one intention trigger exceeds the set task threshold, the intention trigger is selected into the candidate reply of the robot; when the confidence coefficient of the similarity of one boring knowledge point exceeds the set boring threshold value, the candidate reply of the robot is selected;
4) User demand feedback; user feedback comprises satisfaction and transfer manual work, i, the satisfaction is the evaluation of the answer of the robot by the user, the satisfaction is divided into satisfaction, unsatisfactory content and unmatched answer in the system, and the meanings represented by the satisfaction are respectively: satisfaction: the robot answers correctly, and the reply content solves the problem of the user; content dissatisfaction: the robot answers correctly, but the answer content itself may not completely solve the user problem; answer mismatch: the robot matches the wrong knowledge points and answers questions; in the question-answer satisfaction statistics, checking feedback results of the terminal user; ii. The manual transfer refers to that when the robot cannot answer the user message, the robot is transferred to manual customer service for processing, and your manual customer service team can use the instant communication function of the system and can also use a customer service workbench of a third party;
5) Learning new knowledge points; the method comprises the steps of checking new knowledge points and mining the knowledge points; i. the new knowledge points are audited, the chat logs of the users can be periodically imported into the system, the corpus can be uploaded, the denoising processing is automatically carried out, the chat corpus is clear and duplicate-removed, the chat corpus is recorded as an unverified problem, and the chat corpus is arranged in an inverted order according to the date of addition; supporting automatic clustering, merging recommended problems into existing knowledge points, adding new knowledge points, deleting problems, searching existing knowledge points and merging; ii. The system can search a large amount of historical corpus through the knowledge point clustering function and aggregate similar problems, so that the purpose of constructing knowledge points is achieved; before using the knowledge point clustering function, the corpus needs to be uploaded;
6) Personalized questions and answers (property groups and replies); the attribute group and the replies are the centralized management of knowledge points and replies thereof associated with different user attributes, and after the knowledge points are triggered by the messages sent by a plurality of users, different answers are given according to the different user attributes; the attribute group and reply module integrates three elements of an attribute group (user attribute label combination), a knowledge point and a reply into one management view, so that the maintenance can be more convenient and efficient;
7) Training a model; each module provides a training and publishing interface, can specify a model used on line, is executed by a single process for training and evaluating, does not influence the online service process, monitors the success of the training and the model by using a cross-platform watch to monitor the local folder change notification when the single machine is deployed, and carries out the loading by using an NFS synchronous model file and a redis notification when the multiple machines are deployed;
through steps S301 to S303, relative to the problem that the efficiency and accuracy of acquiring and processing military domain information by people in related operations are low, the embodiment of the application constructs a military domain knowledge base, analyzes user intention by acquiring user input, determines answer content based on the pre-constructed military domain knowledge base, and improves the efficiency and accuracy of acquiring and processing the military domain information by people. In addition, the embodiment of the application can realize voice recognition in a specific noise environment, further adapt to various business application scenes and realize quick and accurate conversion from voice to text; on the basis of voice recognition, the embodiment of the application also converts spoken language with specific context into query sentences which can be recognized by a system, and accurately judges the intention of the field problem through methods such as machine learning, semantic understanding and the like; the embodiment of the application establishes a multi-engine question-answer strategy, rapidly inquires and searches a knowledge base through a plurality of modes such as knowledge map question-answer, reading and understanding question-answer, question-answer and library question-answer, and the like, sorts, screens and merges query results, and displays the question-answer results in a plurality of visual modes; when the intention of the problem is fuzzy, the focus requirement of the user is guided through multiple rounds of conversations, related problems and results are recommended, and the experience degree of the user and the accuracy of information acquisition are greatly improved.
The embodiment of the application also provides a platform architecture of the military domain knowledge base, and fig. 5 is a platform architecture diagram of the military domain knowledge base according to the fourth embodiment of the application, as shown in fig. 5, the knowledge base platform mainly comprises three parts of a data layer, a processing layer and an application display layer, wherein the data layer is responsible for data access and data acquisition, provides standardized acquisition and access tools, and leads to different sources and types of data, and mainly comprises military service data, open source data, a third party database and the like; aiming at the problems of scattered data, inconsistent formats and the like of each data source, the processing layer gathers and standardizes multi-source heterogeneous data, and processes the data resources through technical means such as standardized processing and cleaning technology, data storage, data archiving, data synchronous updating and the like; the application presentation layer provides data integration service, knowledge complex retrieval, data content management, data peripheral management and data interface management.
The embodiment of the application also provides the overall architecture of a military field question-answering system, fig. 6 is an overall architecture diagram of the military field question-answering system according to the fifth embodiment of the application, and as shown in fig. 6, the system comprises a user terminal, a man-machine interaction system and an intelligent question-answering server, wherein the man-machine interaction system provides voice recognition, voice synthesis and voice analysis, the intelligent question-answering service provides question-answering semantic understanding, intelligent question-answering and intelligent document management, the user terminal and the man-machine interaction system can conduct data interaction, and the man-machine interaction system can conduct data interaction with the intelligent question-answering server; the intelligent question-answering semantic understanding comprises intention recognition, entity extraction and NL2SQL, the intelligent question-answering comprises FAQ question-answering, map question-answering and MRC, and the intelligent document management comprises data access, knowledge extraction and map construction.
The embodiment of the application also provides a construction process of the military domain knowledge base, the military domain knowledge base is applied to the question-answering method based on the military domain knowledge graph, fig. 7 is a flowchart of the construction process of the military domain knowledge base according to the sixth embodiment of the application, as shown in fig. 7, the flowchart includes the following steps:
step S701, leading and integrating data, and integrating the data by a data processing engine, wherein the data comprises service data, open source data and third party data;
for example, under the support of a data processing engine, data format conversion, cleaning, correction and the like are implemented, association analysis is carried out on structured, semi-structured and unstructured data, safe and efficient connection of various business system data is realized, and integrated data is efficiently managed, wherein the data management system comprises three modules of data access, data integration and data resource management: a. the data access comprises data acquisition, data processing and data distribution; i. the data acquisition utilizes a data source interface component to access various business data such as sea-related image video data, government information data, social resource data, military business data and the like, and the data source is divided into three types of real-time data, quasi-real-time data and basic data according to the difference of sources and transmission modes of the data, so that various transmission modes including TCP, UDP, web Service, FTP, messages, streaming media and mobile storage are supported; the data access is connected to various external data sources through a data source interface component, and comprises but is not limited to databases such as Mysql, oracle, SQL Server and the like, message queues such as RocketMQ, kafka and the like, file transfer services such as FTP, SFTP and the like, streaming media such as RTSP, RTMP and the like are implemented, data acquired from the external data sources are respectively written into different file systems according to different data types, and the data are respectively written into the different file systems according to the different data types, wherein the data comprises but is not limited to databases such as Mysql, oracle, SQL Server and the like, distributed storage systems such as HDFS, swift and the like, data warehouse Hive, HBase and the like, and the file systems form a data original library; ii. The data processing is carried out, the accessed heterogeneous data is converted according to a certain standard specification, the analysis of the data and the call of upper application service are convenient in the later period, the data is filtered, the data information is distinguished and separated through the configuration of the filtering rule, the data which does not meet the service requirement and the data standard rule is filtered, and the error, redundancy and junk data information are removed; secondly, setting corresponding judging data repetition rules for various service scenes, including merging and clearing strategies, identifying repeated data, and merging or clearing the repeated data; setting data conversion rules according to specific service requirements, and unifying the same type of data of different service systems in format; finally, setting data verification rules including but not limited to null value verification, date format verification, number length verification and the like, and verifying the correctness, the integrity and the consistency of the data to generate data meeting the standard and the quality requirement; iii, data distribution, namely distributing the processed data to a specified basic library, a subject library and a subject library according to service application requirements; b. the data integration is realized by constructing a data resource system, leading and converging various social resource data, sea government data, military service data and the like, realizing the integrated management and analysis of the data, and mainly solving the problems of the distribution and the isomerism of the data through the data exchange among the applications; c. the data resource management module performs functions of report visualization, data source configuration, asset management, operation and maintenance service and the like on various data; i. the report visualization provides data analysis and display services based on a statistical chart, can display different visualization styles of data according to different data display and analysis requirements, supports various styles including an indication card, a metering chart, a pie chart, a histogram, a split chart, a word cloud, a radar chart, a funnel chart and the like, supports a user to customize a special visualization component for display, ii, data source configuration, and provides a visualized data source management service based on a web technology for a data analysis task, so that new, modification, deletion and query operations can be conveniently and rapidly performed on a data source to be analyzed, and the stable operation of a resource library, a special question library, a main question library and an external convergence library can be ensured through an online database maintenance mode; iii, asset management, which is to support the inventory of data assets taking files as carriers, such as photos, documents, drawings, videos, digital copyrights and the like, wherein the inventory content comprises data quality, data size, data flow, data update rate, data integrity rate, data use rate and the like, and the inventory content is submitted in a data asset inventory report form, which is to support the quick query of the data assets through attribute information (owners, users, names, types) and the like of the data assets, and to support fuzzy query; iv, operation and maintenance service, data operation and maintenance, including addition, deletion, modification, update, migration, backup and the like of data, and data warehousing: performing operations such as warehousing configuration, batch warehousing, manual warehousing and the like on data, and updating the data: and performing operations such as updating policy customization, manual updating operation, automatic updating operation, updating backtracking inquiry and the like on various data, and migrating the data: and performing migration policy customization, manual migration operation, automatic migration operation, migration backtracking inquiry and other operations on various data, and backing up the data: and performing operations such as backup strategy customization, manual backup operation, automatic backup operation, backup backtracking inquiry and the like on various data, and deleting the data: the operations of single deletion, batch deletion, record deletion and the like of the stored data can be performed according to various modes such as layers, elements, tenses, ranges and the like, and the data is queried: keyword inquiry, combined inquiry, year inquiry, area inquiry, theme inquiry, statistics and summarization of inquiry results and the like are carried out on the data which are put in storage, and cataloging inquiry can be supported under the condition of having data cataloging;
Step S702, extracting target entity, entity attribute and naming relation among entities from the data, fusing and disambiguating the extracted knowledge, and storing the knowledge in the knowledge base of military field;
for example, knowledge extraction comprises entity extraction, relation extraction and attribute extraction from source data, and fusion comprises frame matching, entity alignment, conflict detection and resolution, entity disambiguation, entity linking and knowledge merging; knowledge extraction: i. the entity extraction module supports extraction of specific entities and entity attributes (perfecting description of the entities) from various data (including structured, semi-structured and unstructured data), and realizes automatic word segmentation, part-of-speech tagging, keyword extraction and the like of Chinese and English texts; ii. The relation extraction is to extract a named relation among entities under the condition that entity pairs in a text are known, and the extracted entity pairs and the relation are normalized and expressed in a form of a triplet (arg 1, relation, arg 2), wherein arg1 and arg2 represent semantic relation among the two entities, and the relation extraction comprises relation extraction facing a structured text, relation extraction facing a semi-structured text and relation extraction facing an unstructured text: the relation extraction for structured data adopts DM (Direct Mapping) issued by RDB2RDF working group of W3C for relation extraction, the relation extraction for semi-structured data adopts an open information extraction method, and the relation extraction for unstructured text adopts a relation extraction method based on weak supervision learning; knowledge fusion: i. frame matching, wherein the knowledge frame mainly comprises concepts, attributes, relations and constraints among the concepts, the attributes and the relations, and the frame matching method adopts elements in different knowledge bases, such as whether the concept "astronaut" matches the "astronaut"; ii. Entity alignment is a process of judging whether two entities in the same or different knowledge bases represent the same physical object or not, wherein the knowledge between knowledge bases is shared through the entities of the same concept described by the alignment; firstly, identifying conflict contents of the same entity in different data sources by finding different entities with the same attribute and relationship, then, processing the conflict, and determining proper content selection according to a certain strategy, wherein the conflict processing strategy comprises conflict omission, conflict avoidance and conflict resolution; iv, disambiguation of the entity, solving the problem of name ambiguity widely existing in text information, and disambiguating the entity extraction result to obtain disambiguated entity information; v, entity linking, namely providing an entity linking technology based on a domain map in the project, firstly fusing the similarity measurement characteristics of aspects such as literal similarity, text similarity, synonym set, character inclusion and the like, and secondly incorporating the relevance of the context entity in the text in the map into the candidate characteristics of the link to realize the efficient linking of the domain entity; vi, knowledge merging, namely obtaining knowledge input from a third-party knowledge base product or existing structured data, merging the external knowledge bases and the relational database, and constructing a new knowledge graph;
It should be noted that, the knowledge base platform storage management can store the knowledge base by adopting different types of storage modes according to the characteristics of the data content and support the fast query and search of the knowledge base; i. the knowledge base storage function supports three modes of relational storage, memory storage and graph data storage, and can carry out encryption management on stored data; ii. Knowledge retrieval including full text retrieval, multidimensional retrieval, custom retrieval, and knowledge reasoning retrieval; maintaining knowledge data, namely performing adding and deleting work on the data, reducing useless information and supplementing useful information; the knowledge platform can update and manage, associate and fuse new guide collected data resources, combine system record data, realize knowledge automation and semi-automatic update through methods such as entity link, associate and fuse, ensure knowledge base dynamic fresh-keeping, support manual management maintenance, verification, and modification at the same time; i. the new data source association provides a new data source association function, and the system can associate and fuse the data resources collected by the new lead and record data by combining with the system; ii. Knowledge base updates include automatic and semi-automatic updates; manual management, functions including manual maintenance, manual verification, and manual modification functions: the manual maintenance function, the system provides knowledge adding rules, modifying rules and deleting rules, and a user can maintain a database, a knowledge base and the like of the system according to the rules; through MySQL clients such as navcat, SQLyog and the like, an authorized user can perform adding, deleting, modifying and checking operations on data, and functions such as updating, adding, deleting, inquiring and the like of a knowledge base are provided; and a manual verification function, wherein a user can perform matching verification and the like on the knowledge base model.
Through steps S701 to S702, according to the service requirements and the data actual conditions, the embodiment of the application constructs the professional knowledge base in the field, realizes the deep integration of the data, fuses the isolated fragments of the data into the panoramic jigsaw, and effectively improves the data resource utilization capability.
The embodiment of the application also provides a knowledge fusion architecture diagram, fig. 8 is a knowledge fusion architecture diagram of a military domain knowledge base according to a seventh embodiment of the application, and as shown in fig. 8, knowledge fusion comprises fusion in a vertical direction and fusion in a horizontal direction; the fusion in the vertical direction refers to the fusion of a (higher) layer general body and a (lower) layer domain body or instance data; the fusion in the horizontal direction refers to the fusion of knowledge maps of the same level, so as to realize complementation of example data; the knowledge fusion functional module in the embodiment of the application comprises frame matching, entity alignment, conflict detection and resolution, entity disambiguation, entity linking and knowledge merging.
In addition, in combination with the question-answering method based on the military domain knowledge graph in the above embodiment, the embodiment of the application can be realized by providing a storage medium. The storage medium has a computer program stored thereon; the computer program, when executed by the processor, implements any of the question-answering methods of the above embodiments based on military domain knowledge maps.
In one embodiment, a computer device is provided, which may be a terminal. The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program, when executed by the processor, implements a question-answering method based on a military domain knowledge graph. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, can also be keys, a track ball or a touch pad arranged on the shell of the computer equipment, and can also be an external keyboard, a touch pad or a mouse and the like.
In one embodiment, fig. 9 is a schematic diagram of an internal structure of an electronic device according to an eighth embodiment of the present application, as shown in fig. 9, and an electronic device, which may be a server, is provided, and an internal structure diagram thereof may be as shown in fig. 9. The electronic device includes a processor, a network interface, an internal memory, and a non-volatile memory connected by an internal bus, where the non-volatile memory stores an operating system, computer programs, and a database. The processor is used for providing computing and control capability, the network interface is used for communicating with an external terminal through network connection, the internal memory is used for providing an environment for the operation of an operating system and a computer program, the computer program is executed by the processor to realize a question-answering method based on a military domain knowledge graph, and the database is used for storing data.
It will be appreciated by those skilled in the art that the structure shown in fig. 9 is merely a block diagram of a portion of the structure associated with the present application and is not limiting of the electronic device to which the present application is applied, and that a particular electronic device may include more or fewer components than shown, or may combine certain components, or have a different arrangement of components.
Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the various embodiments provided herein may include non-volatile and/or volatile memory. The nonvolatile memory can include Read Only Memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), memory bus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), among others.
The above examples merely represent a few embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the invention. It should be noted that it would be apparent to those skilled in the art that various modifications and improvements could be made without departing from the spirit of the present application, which would be within the scope of the present application. Accordingly, the scope of protection of the present application is to be determined by the claims appended hereto.
Claims (10)
1. The question answering method based on the military domain knowledge graph is characterized by comprising the following steps of:
acquiring a user voice question, and performing voice recognition on the question to generate a corresponding text, or acquiring a user text question to obtain a corresponding text;
determining the semantics of the text, obtaining the intention of a user, and determining answer content based on a pre-constructed military domain knowledge base according to the intention.
2. The method of claim 1, wherein determining the text semantics comprises:
extracting semantic elements from the text, and judging the intention of a question based on a preset rule or model according to the extracted semantic elements;
And according to the intention discrimination result, converting the question semantic elements into system query sentences so as to query the answer content in the military domain knowledge base.
3. The method of claim 2, wherein, in the event that the user intent cannot be determined, prior to said converting question semantic elements to system query statements, the method comprises:
and distributing the intention to a plurality of task scenes to develop a multi-round dialogue, wherein in one multi-round process, if the user expresses the intention of another task scene, the user enters into another task scene, and the plurality of task scenes share word slot information.
4. The method of claim 1, wherein the determining answer content comprises:
determining a query result through a multi-engine question-answer strategy, sequencing, screening and merging the query result, determining answer content of a user question, and displaying the answer content in a visual form;
wherein, the multi-engine question-answering strategy comprises: and based on knowledge graph questions and answers, reading and understanding the questions and answers, and the questions and answers are versus library questions and answers, performing retrieval in the military domain knowledge base to obtain the query result.
5. The method of claim 1, wherein the constructing of the military domain knowledge base comprises:
the method comprises the steps of leading and integrating data, and integrating the data through a data processing engine, wherein the data comprises business data, open source data and third party data;
and extracting the target entity, the entity attribute and the naming relation among the entities from the data, fusing and disambiguating the extracted knowledge, and storing the knowledge in the military domain knowledge base.
6. A question-answering system based on a military domain knowledge graph, the system comprising:
the acquisition module is used for acquiring a user voice question and carrying out voice recognition on the question to generate a corresponding text, or acquiring a user text question to obtain the corresponding text;
and the determining module is used for determining the semantics of the text, obtaining the intention of the user, and determining the answer content based on a pre-constructed military domain knowledge base according to the intention.
7. The system of claim 6, wherein in the determining module, the determining the text semantics comprises:
extracting semantic elements from the text, and judging the intention of a question based on a preset rule or model according to the extracted semantic elements; and according to the intention discrimination result, converting the question semantic elements into system query sentences so as to query the answer content in the military domain knowledge base.
8. The system of claim 7, wherein the determination module is further configured to, in the event that the user intent cannot be determined, prior to converting the question semantic elements into the system query statement:
and distributing the intention to a plurality of task scenes to develop a multi-round dialogue, wherein in one multi-round process, if the user expresses the intention of another task scene, the user enters into another task scene, and the plurality of task scenes share word slot information.
9. An electronic device comprising a memory and a processor, wherein the memory has stored therein a computer program, the processor being arranged to run the computer program to perform the military domain knowledge-graph based question-answering method of any one of claims 1 to 5.
10. A storage medium having a computer program stored therein, wherein the computer program is configured to perform the military domain knowledge-graph based question-answering method of any one of claims 1 to 5 when run.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310281351.3A CN116414961A (en) | 2023-03-17 | 2023-03-17 | Question-answering method and system based on military domain knowledge graph |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310281351.3A CN116414961A (en) | 2023-03-17 | 2023-03-17 | Question-answering method and system based on military domain knowledge graph |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116414961A true CN116414961A (en) | 2023-07-11 |
Family
ID=87052470
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310281351.3A Pending CN116414961A (en) | 2023-03-17 | 2023-03-17 | Question-answering method and system based on military domain knowledge graph |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116414961A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115497477A (en) * | 2022-09-09 | 2022-12-20 | 平安科技（深圳）有限公司 | Voice interaction method, voice interaction device, electronic equipment and storage medium |
| CN117370506A (en) * | 2023-07-21 | 2024-01-09 | 中图科信数智技术(北京)有限公司 | Agricultural intelligent question-answering method and system supporting multi-mode and multi-round dialogue |
| CN117786071A (en) * | 2023-12-19 | 2024-03-29 | 暗物质(北京)智能科技有限公司 | Question and answer method and device based on service, electronic equipment and storage medium |
-
2023
- 2023-03-17 CN CN202310281351.3A patent/CN116414961A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115497477A (en) * | 2022-09-09 | 2022-12-20 | 平安科技（深圳）有限公司 | Voice interaction method, voice interaction device, electronic equipment and storage medium |
| CN117370506A (en) * | 2023-07-21 | 2024-01-09 | 中图科信数智技术(北京)有限公司 | Agricultural intelligent question-answering method and system supporting multi-mode and multi-round dialogue |
| CN117786071A (en) * | 2023-12-19 | 2024-03-29 | 暗物质(北京)智能科技有限公司 | Question and answer method and device based on service, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3819792A2 (en) | Method, apparatus, device, and storage medium for intention recommendation | |
| CN111753060B (en) | Information retrieval method, apparatus, device and computer readable storage medium | |
| CN110765244B (en) | Method, device, computer equipment and storage medium for obtaining answering operation | |
| CN110888990B (en) | Text recommendation method, device, equipment and medium | |
| WO2021114810A1 (en) | Graph structure-based official document recommendation method, apparatus, computer device, and medium | |
| CN116414961A (en) | Question-answering method and system based on military domain knowledge graph | |
| CN109947902B (en) | Data query method and device and readable medium | |
| US11023503B2 (en) | Suggesting text in an electronic document | |
| CN111557000B (en) | Accuracy Determination for Media | |
| CN111723256A (en) | Government affair user portrait construction method and system based on information resource library | |
| KR20200145299A (en) | Intelligent recruitment support platform based on online interview video analysis and social media information analysis | |
| CN118277588B (en) | Query request processing method, electronic device and storage medium | |
| US20180365324A1 (en) | Method of data organization and data searching for use in constructing evidence-based beliefs | |
| CN114996549A (en) | Intelligent tracking method and system based on active object information mining | |
| CN113870998A (en) | Interrogation method, device, electronic equipment and storage medium | |
| CN117591546A (en) | Query statement generation method and device, electronic equipment and storage medium | |
| CN117708270A (en) | Enterprise data query method, device, equipment and storage medium | |
| CN111931034A (en) | Data searching method, device, equipment and storage medium | |
| CN116361428A (en) | Question-answer recall method, device and storage medium | |
| CN114860883A (en) | Intelligent question and answer method and system | |
| CN115269862A (en) | Electric power question-answering and visualization system based on knowledge graph | |
| CN113449094A (en) | Corpus obtaining method and device, electronic equipment and storage medium | |
| CN111949781A (en) | Intelligent interaction method and device based on natural sentence syntactic analysis | |
| CN117931858B (en) | Data query method, device, computer equipment and storage medium | |
| CN113987146B (en) | Dedicated intelligent question-answering system of electric power intranet |
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 |