CN114428851A - Abstract generation method and device, electronic equipment and storage medium - Google Patents
Abstract generation method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN114428851A CN114428851A CN202111642387.7A CN202111642387A CN114428851A CN 114428851 A CN114428851 A CN 114428851A CN 202111642387 A CN202111642387 A CN 202111642387A CN 114428851 A CN114428851 A CN 114428851A
- Authority
- CN
- China
- Prior art keywords
- dialogue data
- keyword
- layer
- inputting
- data
- 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 51
- 239000013598 vector Substances 0.000 claims description 37
- 230000011218 segmentation Effects 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 9
- 238000012163 sequencing technique Methods 0.000 claims description 6
- 239000000126 substance Substances 0.000 claims 1
- 230000008569 process Effects 0.000 abstract description 4
- 238000013473 artificial intelligence Methods 0.000 abstract description 3
- 238000013135 deep learning Methods 0.000 abstract description 2
- 238000003058 natural language processing Methods 0.000 abstract description 2
- 238000004891 communication Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 9
- 238000012545 processing Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000010276 construction Methods 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
- 239000000284 extract Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual 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
- G06F16/34—Browsing; Visualisation therefor
- G06F16/345—Summarisation for human users
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Abstract
The disclosure provides a method and a device for generating an abstract, electronic equipment and a storage medium, and relates to the technical field of artificial intelligence, in particular to the technical field of deep learning and natural language processing. The specific implementation scheme is as follows: the method comprises the steps of obtaining dialogue data to be processed, inputting the dialogue data into a summary generation model, extracting key words in the dialogue data through the summary generation model, and generating summary information containing the key words, so that the summary information containing the key words is generated for the dialogue data through the summary generation model in the process of generating the dialogue summary, the key information of the dialogue data is contained in the summary information, the situation that the subject deviation occurs in the summary is avoided, and the accuracy of the generated summary is improved.
Description
Technical Field
The present disclosure relates to the field of artificial intelligence technologies, and in particular, to the field of deep learning and natural language processing technologies, and in particular, to a method and an apparatus for generating an abstract, an electronic device, and a storage medium.
Background
With the development of society and the progress of communication technology, the amount of dialogue data is increasing, and different forms of dialogue data are simultaneously generated, and it is very important to extract important information from the various forms of dialogue data.
Disclosure of Invention
The disclosure provides a summary generation method, a device, an electronic device and a storage medium.
According to an aspect of the present disclosure, there is provided a digest generation method including: acquiring dialogue data to be processed; inputting the dialogue data into a summary generation model so as to generate summary information containing keywords of the dialogue data through the summary generation model.
According to another aspect of the present disclosure, there is provided a digest generation apparatus including: the acquisition module is used for acquiring the dialogue data to be processed; and the generation module is used for inputting the dialogue data into a summary generation model so as to generate summary information containing the keywords of the dialogue data through the summary generation model.
According to another aspect of the present disclosure, there is provided an electronic device including: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the summary generation method of the present disclosure.
According to another aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute a digest generation method of an electronic device as claimed in an embodiment of the present disclosure.
According to another aspect of the present disclosure, there is provided a computer program product comprising a computer program which, when executed by a processor, implements the summary generation method of the present disclosure.
One embodiment in the above application has the following advantages or benefits:
the method comprises the steps of obtaining dialogue data to be processed, inputting the dialogue data into a summary generation model, extracting key words in the dialogue data through the summary generation model, and generating summary information containing the key words, so that the summary information containing the key words is generated for the dialogue data through the summary generation model in the process of generating the dialogue summary, the key information of the dialogue data is contained in the summary information, the situation that the subject deviation occurs in the summary is avoided, and the accuracy of the generated summary is improved.
It should be understood that the statements in this section do not necessarily identify key or critical features of the embodiments of the present disclosure, nor do they limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description.
Drawings
The drawings are included to provide a better understanding of the present solution and are not to be construed as limiting the present disclosure. Wherein:
fig. 1 is a schematic flow chart diagram of a digest generation method according to an embodiment of the present disclosure;
fig. 2 is a schematic flow chart diagram of another digest generation method provided in accordance with an embodiment of the present disclosure;
FIG. 3 is a flow chart diagram of another summary generation method provided in accordance with one embodiment of the present disclosure;
FIG. 4 is a schematic flow chart of generating summary information including keywords according to another embodiment of the present disclosure;
fig. 5 is a schematic structural diagram of a summary generation apparatus according to an embodiment of the present disclosure;
fig. 6 is a schematic structural diagram of a summary generation apparatus according to another embodiment of the present disclosure;
fig. 7 is a block diagram of an electronic device for implementing the summary generation method of the embodiments of the present disclosure.
Detailed Description
Exemplary embodiments of the present disclosure are described below with reference to the accompanying drawings, in which various details of the embodiments of the disclosure are included to assist understanding, and which are to be considered as merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the present disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
A digest generation method, apparatus, electronic device, and storage medium of the embodiments of the present disclosure are described below with reference to the accompanying drawings.
Fig. 1 is a flowchart illustrating a digest generation method according to an embodiment of the present disclosure.
It should be noted that an execution subject of the digest generation method in this embodiment is a digest generation apparatus, the digest generation apparatus may be implemented in a software and/or hardware manner, the digest generation apparatus in this embodiment may be configured in an electronic device, the electronic device in this embodiment may include a server, and the embodiment does not specifically limit the electronic device.
As shown in fig. 1, the digest generation method may include:
In some embodiments, the conversation data may be, but is not limited to, conference data, mail data, chatting data, discussion data, and debate data, and the embodiment is not particularly limited thereto.
In some embodiments, after inputting the dialog data into the summary generation model, the summary generation model may perform keyword extraction from the dialog data to extract keywords in the dialog data, and then generate summary information containing the keywords based on the keywords and the dialog data. Therefore, the keyword abstract information containing the dialogue data is generated through an abstract generation model, so that the keyword abstract information contains the keyword information of the dialogue data, the situation that the abstract has theme deviation is avoided, and the accuracy of the generated abstract is improved.
In other embodiments, after inputting the dialog data into the abstract generation model, the abstract generation model may generate candidate abstract information of the dialog data, perform keyword extraction on the dialog data to obtain keywords in the dialog data, adjust the candidate abstract information according to the keywords to generate target abstract information containing the keywords, and output the target abstract information.
The summary information may include keywords and corresponding expansion information for the keywords.
The invention provides a summary generation method, which comprises the steps of obtaining dialogue data to be processed, inputting the dialogue data into a summary generation model, extracting key words in the dialogue data through the summary generation model, and generating summary information containing the key words.
Based on the above embodiments, in order to enable the summary generation model to accurately generate the summary information containing the keywords, in some embodiments, the summary generation model includes: a keyword recognition layer and a summary generation layer, the summary generation method of this embodiment is further described below with reference to fig. 2, as shown in fig. 2.
In some embodiments, the keyword recognition layer may perform keyword extraction on the dialog data to obtain keywords in the dialog data.
As an exemplary embodiment, the keyword recognition layer may perform word segmentation on the dialogue data to obtain each segmented word in the dialogue data, determine a semantic feature vector of each segmented word, and determine whether each segmented word is a keyword of the dialogue data based on the semantic feature vector of each segmented word.
In some embodiments, one implementation of inputting the dialog data and the keyword into the summary generation layer to generate the summary information including the keyword may be to determine a key sentence in the dialog data according to the dialog data and the keyword, input the keyword and the key sentence into the summary generation layer to generate the summary information including the keyword, thereby improving the generation efficiency of the summary information.
Specifically, after the abstract generation layer acquires the dialogue data and the keywords, a plurality of sentences in the dialogue data are determined, the semantic relevance between the sentences and the keywords is determined for each sentence, the sentences are determined as the key sentences in the dialogue data when the semantic relevance is greater than or equal to a preset relevance threshold, and the keywords and the key sentences are input to the abstract generation layer to generate abstract information containing the keywords.
In the present embodiment, a keyword in the dialogue data is determined by the keyword recognition layer of the summary generation model, and the keyword and the dialogue data are input into the summary generation layer of the summary generation model to generate summary information containing the keyword. Therefore, the keywords in the dialogue data are preferentially extracted, so that the abstract generation model can more effectively identify the key information of the dialogue data and perform abstract generation based on the key information, and the efficiency of generating the abstract information containing the keywords by the abstract generation model is improved.
Based on any one of the above embodiments, in order to further improve the efficiency of generating the summary information by the summary generation model, the summary generation model of this embodiment may include an input layer, an encoding layer, a keyword recognition layer, and a summary generation layer, which are connected in sequence, and the summary generation method of this embodiment is further described below with reference to fig. 3. As shown in fig. 3, may include:
Specifically, the input layer performs word segmentation on the dialogue data to obtain each word segmentation of the dialogue data, and performs vector representation on each word segmentation to obtain a word vector of each word segmentation in the dialogue data.
The encoding layer extracts semantic features of each participle according to a word vector of each participle in the dialogue data and outputs a semantic feature vector representing the semantic features of each participle, wherein the semantic features of any participle can comprise semantic information of the participle and can also comprise context semantic information related to the participle in the dialogue data.
In some embodiments, in order to accurately determine the keywords in the dialog data, the keyword recognition layer includes a probability determination sub-layer (not shown in fig. 2) and a keyword determination sub-layer (not shown in fig. 2), and one possible implementation manner of inputting the semantic feature vector of each segmented word into the keyword recognition layer to obtain the keywords in the dialog data is as follows: inputting the semantic feature vector of each participle into a probability determination sublayer to obtain the probability that each participle is a keyword of the dialogue data; the probabilities that the respective segmented words are the keywords of the dialogue data are input to the keyword determination sublayer to determine the keywords in the dialogue data.
In different application scenarios, the keyword determination sub-layer determines the keywords in the dialogue data in different ways by combining the probabilities that the respective participles are the keywords in the dialogue data, and an exemplary embodiment is as follows:
as an exemplary embodiment, the respective participles are sorted according to the order from the greater probability to the smaller probability to obtain a sorting result; and taking the word segmentation ranked at the top N in the ranking result as a keyword in the dialogue data, wherein N is an integer greater than or equal to 1.
The N is set according to actual service requirements, and for example, the N may be 1, 2, or 3.
As another exemplary embodiment, for each participle, comparing the probability that the participle is a keyword of the dialogue data with a probability threshold value preset in the keyword determination sublayer; and determining the participles with the probability greater than or equal to the probability threshold value as the keywords in the dialogue data.
The probability threshold is set to be high, the accuracy is high, and the recall rate is correspondingly reduced. If the probability threshold is set to be low, the accuracy is low, the recall rate is high, and the threshold can be set according to needs, for example, the probability threshold can be set to be 0.6.
In this embodiment, the word vector of each participle in the dialog data is determined by the input layer of the abstract generation model, the word vector of each participle is input into the encoding layer to obtain the semantic feature vector of each participle, the semantic feature vector of each participle is input into the keyword recognition layer to obtain the keyword in the dialog data, and the dialog data and the keyword are input into the abstract generation layer of the abstract generation model to generate abstract information containing the keyword. Therefore, the keywords in the dialogue data are preferentially extracted through the keyword identification layer in the abstract generation model, so that the abstract generation model can effectively identify the key information of the dialogue data and carry out abstract generation based on the key information, the efficiency of generating the abstract information containing the keywords by the abstract generation model is improved, and the problem of abstract theme deviation can be effectively avoided.
In order to clearly understand the present disclosure, the summary generation method of this example is described below with reference to fig. 4. In fig. 4, the example is given by taking the conversation data as the conversation data in the conference process, and the conversation data is "the subject of the today's conference is a patent application". As shown in fig. 4, specifically, an input layer in the abstract generation model performs word segmentation on the dialogue data first, and obtains a word segmentation result of the dialogue data, where the word segmentation result includes each word segmentation of the dialogue data, the word segmentation result of the dialogue data that the topic of the today conference is a patent application includes 7 words, and the 7 words are respectively the topic of the today conference is a patent application. Correspondingly, the input layer carries out vector representation on each participle according to the participle result to obtain a word vector corresponding to each participle, the word vector corresponding to each participle is input into the coding layer to obtain a semantic feature vector of each input word, the semantic feature vector of each input word is input into the keyword identification layer to identify the keywords of the original sentence, for example, the keywords are patent and application, and the keywords and the input words are input into the abstract generation layer to contain the conference abstract of the keywords, for example, the conference abstract is patent application. Therefore, the keywords in the dialogue data are preferentially extracted through the keyword identification layer in the abstract generation model, so that the abstract generation model can effectively identify the key information of the dialogue data and carry out abstract generation based on the key information, the efficiency of generating the abstract information containing the keywords by the abstract generation model is improved, and the problem of abstract theme deviation can be effectively avoided.
Fig. 5 is a schematic structural diagram of a summary generation apparatus according to an embodiment of the present disclosure. As shown in fig. 5, the summary generation apparatus 500 may include: an obtaining module 501 and a generating module 502, wherein:
an obtaining module 501, configured to obtain to-be-processed session data.
The generating module 502 is configured to input the dialog data into the abstract generating model, so as to extract the keywords in the dialog data through the abstract generating model, and generate abstract information including the keywords.
It should be noted that the explanation of the display summary generation method embodiment is also applicable to the display summary generation apparatus in this embodiment, and details are not repeated here.
The invention provides a summary generation device, which is used for acquiring dialogue data to be processed, inputting the dialogue data into a summary generation model, extracting key words in the dialogue data through the summary generation model and generating summary information containing the key words.
In an embodiment of the present disclosure, as shown in fig. 6, the digest generation apparatus may further include: an obtaining module 601, a generating module 602, a first input module 603 and a second input module 604, wherein the generating module 602 comprises an input unit 6021 and a generating unit 6022, wherein the input unit 6021 comprises a first input subunit 60211 and a second input subunit 60212, and the generating unit 6022 comprises a determining subunit 60221 and a generating subunit 60222.
For a detailed description of the obtaining module 601 and the generating module 602, reference is made to the descriptions of the obtaining module 501 and the generating module 502 in the embodiment shown in fig. 5, and the description is not provided here.
In one embodiment of the present disclosure, as shown in fig. 6, the summary generation model includes: a keyword recognition layer and a digest generation layer.
Wherein, the generating module 602 includes:
an input unit 6021, configured to input the dialogue data to the keyword recognition layer to obtain a keyword in the dialogue data.
A generation unit 6022 configured to input the dialogue data and the keyword to the digest generation layer to generate digest information containing the keyword.
In an embodiment of the present disclosure, as shown in fig. 6, the summary generation model further includes an input layer and an encoding layer, and the apparatus further includes:
the first input module 603 is configured to input the dialog data into the input layer to obtain a word vector of each participle in the dialog data.
The second input module 604 is configured to input the word vector of each participle into the coding layer, so as to obtain a semantic feature vector of each participle.
The input unit 6021 is specifically configured to:
and inputting the semantic feature vector of each participle into a keyword recognition layer to obtain keywords in the dialogue data.
In one embodiment of the present disclosure, as shown in fig. 6, the keyword recognition layer includes a probability determination sublayer and a keyword determination sublayer, and the input unit 6021 includes:
a first input subunit 60211, configured to input the semantic feature vector of each participle to the probability determination sublayer, so as to obtain a probability that each participle is a keyword of the dialogue data.
A second input sub-unit 60212 for inputting the probability that each participle is a keyword of the dialogue data to the keyword determination sub-layer to determine the keyword in the dialogue data.
In an embodiment of the present disclosure, as shown in fig. 6, the second input subunit 60212 is specifically configured to:
and sequencing the participles according to the sequence from the large probability to the small probability to obtain a sequencing result.
And taking the word segmentation ranked at the top N in the ranking result as a keyword in the dialogue data, wherein N is an integer greater than or equal to 1.
In an embodiment of the present disclosure, as shown in fig. 6, the second input subunit 60212 is further specifically configured to:
for each participle, the probability that the participle is a keyword of the dialogue data is compared with a probability threshold value preset in the keyword determination sublayer.
And determining the participles with the probability greater than or equal to the probability threshold value as the keywords in the dialogue data.
In one embodiment of the present disclosure, as shown in fig. 6, the green layer unit 6022 includes:
a determining subunit 60221, configured to determine the key sentence in the dialogue data according to the dialogue data and the keyword.
A generation subunit 60222 is configured to input the keyword and the key sentence to the digest generation layer to generate digest information containing the keyword.
In an embodiment of the present disclosure, as shown in fig. 6, the determining subunit 60221 is specifically configured to:
a plurality of statements in the dialog data is determined.
For each sentence, a semantic relatedness between the sentence and the keyword is determined.
And determining the sentence as a key sentence in the dialogue data under the condition that the semantic relevance is greater than or equal to a preset relevance threshold.
The invention provides a summary generation device, which is used for acquiring dialogue data to be processed, inputting the dialogue data into a summary generation model, extracting key words in the dialogue data through the summary generation model and generating summary information containing the key words.
It should be noted that the explanation of the foregoing summary generation method embodiment is also applicable to the summary generation apparatus in this embodiment, and details are not described here.
FIG. 7 shows a schematic block diagram of an example electronic device 700 that may be used to implement embodiments of the present disclosure. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the disclosure described and/or claimed herein.
As shown in fig. 7, the device 700 comprises a computing unit 701, which may perform various suitable actions and processes according to a computer program stored in a Read Only Memory (ROM)702 or a computer program loaded from a storage unit 708 into a Random Access Memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the device 700 can also be stored. The computing unit 701, the ROM 702, and the RAM 703 are connected to each other by a bus 704. An input/output (I/O) interface 705 is also connected to bus 704.
A number of components in the device 700 are connected to the I/O interface 705, including: an input unit 706 such as a keyboard, a mouse, or the like; an output unit 707 such as various types of displays, speakers, and the like; a storage unit 708 such as a magnetic disk, optical disk, or the like; and a communication unit 709 such as a network card, modem, wireless communication transceiver, etc. The communication unit 709 allows the device 700 to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.
Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuitry, Field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), system on a chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs that are executable and/or interpretable on a programmable system including at least one programmable processor, which may be special or general purpose, receiving data and instructions from, and transmitting data and instructions to, a storage system, at least one input device, and at least one output device.
Program code for implementing the methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions/operations specified in the flowchart and/or block diagram to be performed. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine or entirely on the remote machine or server.
In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic, speech, or tactile input.
The systems and techniques described here can be implemented in a computing system that includes a back-end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), Wide Area Networks (WANs), the internet, and blockchain networks.
The computer system may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server may be a cloud server, a server of a distributed system, or a server with a combined blockchain.
It should be understood that various forms of the flows shown above may be used, with steps reordered, added, or deleted. For example, the steps described in the present disclosure may be executed in parallel, sequentially, or in different orders, as long as the desired results of the technical solutions disclosed in the present disclosure can be achieved, and the present disclosure is not limited herein.
The above detailed description should not be construed as limiting the scope of the disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may be made in accordance with design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present disclosure should be included in the scope of protection of the present disclosure.
Claims (19)
1. A summary generation method comprises the following steps:
acquiring dialogue data to be processed;
inputting the dialogue data into a summary generation model so as to generate summary information containing keywords of the dialogue data through the summary generation model.
2. The method of claim 1, wherein the summary generation model comprises: a keyword recognition layer and a summary generation layer;
wherein the inputting the dialogue data into a summary generation model to generate summary information containing keywords of the dialogue data through the summary generation model comprises:
inputting the dialogue data into a keyword recognition layer to obtain keywords in the dialogue data;
inputting the dialogue data and the key words into the abstract generation layer to generate abstract information containing the key words.
3. The method of claim 2, wherein the digest generation model further comprises an input layer, an encoding layer, the method further comprising:
inputting the dialogue data into an input layer to obtain word vectors of each participle in the dialogue data;
inputting the word vector of each participle into the coding layer to obtain a semantic feature vector of each participle;
the inputting the dialogue data into a keyword recognition layer to obtain the keywords in the dialogue data includes:
and inputting the semantic feature vector of each participle into the keyword recognition layer to obtain the keywords in the dialogue data.
4. The method of claim 3, wherein the keyword recognition layer comprises a probability determination sub-layer and a keyword determination sub-layer, and the inputting semantic feature vectors of the respective participles into the keyword recognition layer to obtain keywords in the dialogue data comprises:
inputting the semantic feature vector of each participle into the probability determination sublayer to obtain the probability that each participle is the keyword of the dialogue data;
and inputting the probability that each participle is the keyword of the dialogue data into the keyword determination sublayer so as to determine the keyword in the dialogue data.
5. The method of claim 4, wherein the inputting the probabilities that the respective participles are keywords of the conversational data to the keyword determination sub-layer to determine keywords in the conversational data comprises:
sequencing the word segments according to the sequence from the large probability to the small probability to obtain a sequencing result;
and taking the word segmentation ranked at the top N in the ranking result as the keyword in the dialogue data, wherein N is an integer greater than or equal to 1.
6. The method of claim 4, wherein the inputting the probabilities that the respective participles are keywords of the conversational data to the keyword determination sub-layer to determine keywords in the conversational data comprises:
for each participle, comparing the probability that the participle is the keyword of the dialogue data with a probability threshold value preset in the keyword determination sublayer;
and determining the participles with the probability greater than or equal to a probability threshold value as the keywords in the dialogue data.
7. The method of any of claims 2-6, wherein said inputting the dialog data and the keyword into the summary generation layer to generate summary information containing the keyword comprises:
determining key sentences in the dialogue data according to the dialogue data and the keywords;
and inputting the key words and the key sentences to a summary generation layer to generate summary information containing the key words.
8. The method of claim 7, wherein said determining key sentences in the dialogue data from the dialogue data and the keywords comprises:
determining a plurality of statements in the dialog data;
for each sentence, determining semantic relatedness between the sentence and the keyword;
and determining the sentence as a key sentence in the dialogue data under the condition that the semantic relevance is greater than or equal to a preset relevance threshold.
9. A digest generation apparatus comprising:
the acquisition module is used for acquiring the dialogue data to be processed;
and the generation module is used for inputting the dialogue data into a summary generation model so as to generate summary information containing the keywords of the dialogue data through the summary generation model.
10. The apparatus of claim 9, wherein the summary generation model comprises: a keyword recognition layer and a summary generation layer;
wherein the generating module comprises:
the input unit is used for inputting the dialogue data to a keyword recognition layer so as to obtain keywords in the dialogue data;
a generating unit configured to input the dialogue data and the keyword to the digest generation layer to generate digest information including the keyword.
11. The apparatus of claim 10, wherein the digest generation model further comprises an input layer, an encoding layer, the apparatus further comprising:
the first input module is used for inputting the dialogue data into an input layer so as to obtain word vectors of each participle in the dialogue data;
the second input module is used for inputting the word vectors of all the participles into the coding layer so as to obtain the semantic feature vectors of all the participles;
the input unit is specifically configured to:
and inputting the semantic feature vector of each participle into the keyword recognition layer to obtain the keywords in the dialogue data.
12. The apparatus of claim 11, wherein the keyword recognition layer comprises a probability determination sublayer and a keyword determination sublayer, the input unit comprising:
the first input subunit is used for inputting the semantic feature vector of each participle into the probability determination sublayer so as to obtain the probability that each participle is the keyword of the dialogue data;
and the second input subunit is used for inputting the probability that each participle is the keyword of the dialogue data into the keyword determination sublayer so as to determine the keyword in the dialogue data.
13. The apparatus of claim 12, wherein the second input subunit is specifically configured to:
sequencing the word segments according to the sequence from the large probability to the small probability to obtain a sequencing result;
and taking the word segmentation ranked at the top N in the ranking result as the keyword in the dialogue data, wherein N is an integer greater than or equal to 1.
14. The apparatus of claim 12, wherein the second input subunit is further specifically configured to:
for each participle, comparing the probability that the participle is the keyword of the dialogue data with a probability threshold value preset in the keyword determination sublayer;
and determining the participles with the probability greater than or equal to a probability threshold value as the keywords in the dialogue data.
15. The apparatus of any one of claims 10-14, wherein the layer generating unit comprises:
the determining subunit is used for determining a key statement in the dialogue data according to the dialogue data and the key word;
and the generating subunit is used for inputting the key words and the key sentences into a summary generating layer so as to generate summary information containing the key words.
16. The apparatus of claim 15, wherein the determining subunit is specifically configured to:
determining a plurality of statements in the dialog data;
for each sentence, determining semantic relatedness between the sentence and the keyword;
and determining the sentence as a key sentence in the dialogue data under the condition that the semantic relevance is greater than or equal to a preset relevance threshold.
17. An electronic device, comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-8.
18. A non-transitory computer readable storage medium having stored thereon computer instructions for causing the computer to perform the method of any one of claims 1-8.
19. A computer program product comprising a computer program which, when executed by a processor, carries out the steps of the method of any one of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111642387.7A CN114428851A (en) | 2021-12-29 | 2021-12-29 | Abstract generation method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111642387.7A CN114428851A (en) | 2021-12-29 | 2021-12-29 | Abstract generation method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114428851A true CN114428851A (en) | 2022-05-03 |
Family
ID=81310559
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111642387.7A Pending CN114428851A (en) | 2021-12-29 | 2021-12-29 | Abstract generation method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114428851A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106599148A (en) * | 2016-12-02 | 2017-04-26 | 东软集团股份有限公司 | Method and device for generating abstract |
| CN112364624A (en) * | 2020-11-04 | 2021-02-12 | 重庆邮电大学 | Keyword extraction method based on deep learning language model fusion semantic features |
| CN112528655A (en) * | 2020-12-18 | 2021-03-19 | 北京百度网讯科技有限公司 | Keyword generation method, device, equipment and storage medium |
| CN113190675A (en) * | 2021-05-12 | 2021-07-30 | 平安国际智慧城市科技股份有限公司 | Text abstract generation method and device, computer equipment and storage medium |
| CN113268586A (en) * | 2021-05-21 | 2021-08-17 | 平安科技（深圳）有限公司 | Text abstract generation method, device, equipment and storage medium |
| US20210342552A1 (en) * | 2020-05-01 | 2021-11-04 | International Business Machines Corporation | Natural language text generation from a set of keywords using machine learning and templates |
-
2021
- 2021-12-29 CN CN202111642387.7A patent/CN114428851A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106599148A (en) * | 2016-12-02 | 2017-04-26 | 东软集团股份有限公司 | Method and device for generating abstract |
| US20210342552A1 (en) * | 2020-05-01 | 2021-11-04 | International Business Machines Corporation | Natural language text generation from a set of keywords using machine learning and templates |
| CN112364624A (en) * | 2020-11-04 | 2021-02-12 | 重庆邮电大学 | Keyword extraction method based on deep learning language model fusion semantic features |
| CN112528655A (en) * | 2020-12-18 | 2021-03-19 | 北京百度网讯科技有限公司 | Keyword generation method, device, equipment and storage medium |
| CN113190675A (en) * | 2021-05-12 | 2021-07-30 | 平安国际智慧城市科技股份有限公司 | Text abstract generation method and device, computer equipment and storage medium |
| CN113268586A (en) * | 2021-05-21 | 2021-08-17 | 平安科技（深圳）有限公司 | Text abstract generation method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107301170B (en) | Method and device for segmenting sentences based on artificial intelligence | |
| CN111241819B (en) | Word vector generation method and device and electronic equipment | |
| CN113129870B (en) | Training method, device, equipment and storage medium of speech recognition model | |
| CN112926306B (en) | Text error correction method, device, equipment and storage medium | |
| CN112466288A (en) | Voice recognition method and device, electronic equipment and storage medium | |
| CN114861889B (en) | Deep learning model training method, target object detection method and device | |
| CN112528641A (en) | Method and device for establishing information extraction model, electronic equipment and readable storage medium | |
| CN114724168A (en) | Training method of deep learning model, text recognition method, text recognition device and text recognition equipment | |
| CN113360700A (en) | Method, device, equipment and medium for training image-text retrieval model and image-text retrieval | |
| CN112699237B (en) | Label determination method, device and storage medium | |
| CN112506359A (en) | Method and device for providing candidate long sentences in input method and electronic equipment | |
| CN113553833B (en) | Text error correction method and device and electronic equipment | |
| CN114818736A (en) | Text processing method, chain finger method and device for short text and storage medium | |
| CN114428851A (en) | Abstract generation method and device, electronic equipment and storage medium | |
| CN113408269A (en) | Text emotion analysis method and device | |
| CN113408280A (en) | Negative example construction method, device, equipment and storage medium | |
| CN112528027A (en) | Text classification method, device, equipment, storage medium and program product | |
| CN116244432B (en) | Pre-training method and device for language model and electronic equipment | |
| CN112989797B (en) | Model training and text expansion methods, devices, equipment and storage medium | |
| CN116069914B (en) | Training data generation method, model training method and device | |
| CN113377921B (en) | Method, device, electronic equipment and medium for matching information | |
| CN113377922B (en) | Method, device, electronic equipment and medium for matching information | |
| CN113255332B (en) | Training and text error correction method and device for text error correction model | |
| CN115033701B (en) | Text vector generation model training method, text classification method and related device | |
| CN114491040B (en) | Information mining method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20220503 |
|
| RJ01 | Rejection of invention patent application after publication |