CN112214594A - Text briefing generation method and device, electronic equipment and readable storage medium - Google Patents
Text briefing generation method and device, electronic equipment and readable storage medium Download PDFInfo
- Publication number
- CN112214594A CN112214594A CN202011054465.7A CN202011054465A CN112214594A CN 112214594 A CN112214594 A CN 112214594A CN 202011054465 A CN202011054465 A CN 202011054465A CN 112214594 A CN112214594 A CN 112214594A
- Authority
- CN
- China
- Prior art keywords
- briefing
- label
- vector
- category
- brief report
- 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.)
- Withdrawn
Links
- 238000000034 method Methods 0.000 title claims abstract description 54
- 238000012549 training Methods 0.000 claims abstract description 96
- 239000013598 vector Substances 0.000 claims description 188
- 238000005315 distribution function Methods 0.000 claims description 19
- 230000006870 function Effects 0.000 claims description 17
- 230000009466 transformation Effects 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 7
- 238000012163 sequencing technique Methods 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000005516 engineering process Methods 0.000 abstract description 2
- 206010035664 Pneumonia Diseases 0.000 description 8
- 208000024891 symptom Diseases 0.000 description 8
- 241000711573 Coronaviridae Species 0.000 description 6
- 206010037660 Pyrexia Diseases 0.000 description 6
- 238000013528 artificial neural network Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 238000007726 management method Methods 0.000 description 5
- 238000013135 deep learning Methods 0.000 description 4
- 230000036541 health Effects 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 206010011224 Cough Diseases 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 230000002457 bidirectional effect Effects 0.000 description 2
- 230000036760 body temperature Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 125000004122 cyclic group Chemical group 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000002955 isolation Methods 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000012821 model calculation Methods 0.000 description 2
- 238000003058 natural language processing Methods 0.000 description 2
- 208000027406 nose symptom Diseases 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000008520 organization Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000006467 substitution reaction Methods 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
- 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
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/12—Use of codes for handling textual entities
- G06F40/151—Transformation
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Operations Research (AREA)
- Probability & Statistics with Applications (AREA)
- Algebra (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to an artificial intelligence technology, and discloses a text briefing generation method, which comprises the following steps: the method comprises the steps of training a pre-constructed original label classification network by utilizing a first brief report set and a first category label set to obtain a standard label classification network, executing label classification operation on a second brief report set by utilizing the standard label classification network to obtain a second category label set, summarizing the first category label set and the second category label set to obtain a category label set, dividing the first brief report set and the second brief report set by utilizing the category label set to obtain a category brief report set, generating a brief report template by utilizing the category brief report set, receiving a brief report text to be generated, executing named entity recognition on the brief report text to be generated to obtain an entity set, and executing combined operation on the entity set and the brief report template to obtain a brief report. The invention also provides a text briefing generation device, electronic equipment and a computer readable storage medium. The invention can save the computing resource and improve the accuracy of the generation of the briefing.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a text briefing generation method and device, electronic equipment and a computer readable storage medium.
Background
The brief report is the short information of the transmitted information, also called as 'brief message', and the current text brief report generation method mainly includes an extraction brief report generation method and a generation brief report generation method.
The extraction type briefing generating method comprises the steps of firstly carrying out text vectorization, calculating the similarity between each group of text vectors, storing the similarity into a matrix, and utilizing the matrix to construct a diagram structure to generate a briefing; the generation type briefing generation method inputs vectorized texts into a deep neural network of an encoder-decoder structure, iteratively trains internal parameters of the deep neural network, and then outputs vectors by using the network and decodes the vectors to obtain the briefing.
Both methods can generate the briefs, but the extraction-type briefs generation method only depends on simple mathematical calculation due to vector similarity, the randomness of similarity results is high, the briefs quality needs to be improved, the extraction-type briefs generation method needs to depend on a large amount of training data to train a neural network, the labeling of early-stage training data and the like not only wastes time and labor, but also consumes excessive computer resources to train a model based on a large amount of training data.
Disclosure of Invention
The invention provides a method and a device for generating a text briefing, electronic equipment and a computer readable storage medium, and mainly aims to save computing resources and improve the accuracy of briefing generation.
In order to achieve the above object, the method for generating a text briefing provided by the present invention comprises:
receiving a brief report set, and classifying the brief report set according to a preset proportion to obtain a first brief report set and a second brief report set;
receiving a first class label set corresponding to the first brief report set, and training a pre-constructed original label classification network by using the first brief report set and the first class label set to obtain a standard label classification network;
performing label classification operation on the second brief report set by using the standard label classification network to obtain a second class label set, summarizing the first class label set and the second class label set to obtain a class label set corresponding to the brief report set;
dividing the briefing set into different categories by using the category label set to obtain a category briefing set, and generating a briefing template by using the category briefing set;
receiving a to-be-generated brief report text, executing named entity recognition on the to-be-generated brief report text to obtain an entity set, and executing combined operation on the entity set and the brief report template to obtain a brief report.
Optionally, the training a pre-constructed original label classification network by using the first briefing set and the first class label set to obtain a standard label classification network includes:
respectively executing vectorization operation on the first briefing set and the first category label set by using a vector conversion network in the original label classification network to obtain a first briefing vector set and a first category label vector set;
performing a combination operation on the first briefing vector set and the first class label vector set to obtain a training vector set;
calculating a probability distribution value set of the training vector set by using a probability distribution function in the original label classification network;
and calculating a minimized target value of the probability distribution value set, and when the minimized target value is greater than a preset threshold value, adjusting internal parameters of the original label classification network until the minimized target value is less than or equal to the preset threshold value, so as to obtain the standard label classification network.
Optionally, the calculating a set of probability distribution values of the training vector set by using a probability distribution function in the original label classification network includes:
respectively constructing a first vector set and a second vector set according to the training vector set;
calculating a set of class prototypes for the first set of vectors;
and solving to obtain the probability distribution value set by taking the class prototype set and the second vector set as function parameters of the probability distribution function.
Optionally, the computing a set of class archetypes for the first set of vectors comprises:
calculating a class prototype set of the first vector set by adopting the following method:
wherein, ckClass prototype representing class label k, S (k) being said set of training vectors, fθ(xi) A vector representation function, x, representing the ith training vector in the first vector set in the vector transformation networkiRepresenting the ith training vector, y, in the first set of vectorsiAnd representing a category label vector corresponding to the ith training vector in the first vector set.
Optionally, said solving a set of probability distribution values of the class prototype set and the second vector set using the probability distribution function includes:
solving a set of probability distribution values of the class prototype set and the second vector set by using the following probability distribution functions:
wherein x isjRepresenting the jth training vector, p, in said second set of vectorsθ(y＝k|xj) A probability distribution value, f, representing that the jth training vector in the second vector set belongs to a class label kθ(xj) Representing a vector representation function of a jth training vector of said second set of vectors in said vector transformation network, ckA class prototype representing a class label k.
Optionally, the generating a presentation template by using the category presentation set includes:
performing attribute replacement operation and part-of-speech tagging operation on the category briefing set to obtain a primary briefing set;
and sequencing the sentences of the primary briefing set to obtain the briefing template.
Optionally, the preset ratio is 4: 6.
In order to solve the above problem, the present invention further provides a text briefing generating apparatus, including:
the briefing dividing module is used for receiving a briefing set, classifying the briefing set according to a preset proportion, and obtaining a first briefing set and a second briefing set;
the classification network training module is used for receiving a first class label set corresponding to the first brief report set, and training a pre-constructed original label classification network by using the first brief report set and the first class label set to obtain a standard label classification network;
a category label identification module, configured to perform a label classification operation on the second brief report set by using the standard label classification network to obtain a second category label set, and summarize the first category label set and the second category label set to obtain a category label set corresponding to the brief report set;
the briefing generating module is used for dividing the briefing set into different categories by using the category label set to obtain a category briefing set, generating a briefing template by using the category briefing set, receiving a briefing text to be generated, performing named entity identification on the briefing text to be generated to obtain an entity set, and performing combined operation on the entity set and the briefing template to obtain the briefing.
In order to solve the above problem, the present invention also provides an electronic device, including:
at least one processor; and the number of the first and second groups,
a memory communicatively coupled to the at least one processor; wherein,
the memory stores instructions executable by the at least one processor to cause the at least one processor to perform the method of generating a text presentation described above.
In order to solve the above problem, the present invention further provides a computer-readable storage medium including a storage data area and a storage program area, the storage data area storing created data, the storage program area storing a computer program; wherein the computer program, when executed by a processor, implements a method of generating a text presentation as described in any of the above.
The embodiment of the invention divides the brief report set into two to obtain a first brief report set and a second brief report set, trains a label classification network by utilizing the first brief report set and the corresponding first class label set, performs label classification on the second brief report set by utilizing the trained label classification network to obtain a second class label set, and summarizes the first class label set and the second class label set to obtain the class label set corresponding to the brief report set. Because the embodiment of the invention only uses part of the brief report set training model and classifies the rest brief report sets based on the trained model, compared with relying on the model training using all the brief report training data, the embodiment of the invention not only reduces the label of the brief report training data in the early stage, but also reduces the model calculation amount and saves the computer resources because of the reduction of the training data. Therefore, the method, the device and the computer readable storage medium for generating the text briefing can save computing resources and improve the accuracy of briefing generation.
Drawings
Fig. 1 is a schematic flow chart of a method for generating a text presentation according to an embodiment of the present invention;
fig. 2 is a schematic detailed flowchart of S2 in the text briefing generation method according to an embodiment of the present invention;
fig. 3 is a detailed flowchart of S3 in the text briefing generation method according to an embodiment of the present invention;
fig. 4 is a schematic block diagram of an apparatus for generating a text presentation according to an embodiment of the present invention;
fig. 5 is a schematic internal structural diagram of an electronic device implementing a method for generating a text presentation according to an embodiment of the present invention;
the implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The embodiment of the application provides a text briefing generation method. The execution subject of the text briefing generation method includes, but is not limited to, at least one of electronic devices such as a server and a terminal, which can be configured to execute the method provided by the embodiment of the present application. In other words, the text presentation generation method may be executed by software or hardware installed in the terminal device or the server device, and the software may be a block chain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like.
The invention provides a text briefing generation method. Fig. 1 is a schematic flow chart of a method for generating a text presentation according to an embodiment of the present invention. In this embodiment, the method for generating a text presentation includes:
s1, receiving the briefing sets, and classifying the briefing sets according to a preset proportion to obtain a first briefing set and a second briefing set.
The briefing set is a set of briefing under a specific scene, wherein the specific scene comprises a health field briefing scene, an enterprise analysis briefing scene and the like. Taking a health field briefing scene as an example, under the current broad background of the prevalence of the novel coronavirus pneumonia, how to efficiently and accurately extract brief information from massive and various articles to form a briefing for a user to refer is very important, so that the embodiment of the invention obtains a briefing set related to the novel coronavirus pneumonia from a website, a database and the like by using a crawler or a user collection and arrangement mode and the like.
The preset ratio is generally specified by a user, and in a preferred embodiment of the present invention, the preset ratio is 4:6, wherein the data amount of the first presentation set accounts for 40% of the presentation set, and the data amount of the second presentation set accounts for 60% of the presentation set.
S2, receiving a first category label set corresponding to the first brief report set, and training a pre-constructed original label classification network by using the first brief report set and the first category label set to obtain a standard label classification network.
Preferably, the category labels in the first category label set are determined according to specific scenes, and for a new coronary pneumonia brief report scene, the category labels include category 1 "suspected or confirmed by oneself", category 2 "whether suspected or confirmed by oneself is touched, category 3" fever and discomfort symptom by oneself ", and the like. The first category tag set is generally determined by an expert in a specific scene, and corresponds to the first brief report set in a many-to-one or one-to-one manner, for example, brief reports a exist in the first brief report set, the first sentence of the brief report a is category 1, and the second sentence and the third sentence are category 2.
Further, the original label classification network comprises a vector transformation network and a probability distribution function. In detail, referring to fig. 2, the training of the pre-constructed original tag classification network by using the first briefing set and the first category tag set to obtain the standard tag classification network includes:
s21, executing vectorization operation on the first briefing set and the first category label set by using a vector conversion network in the original label classification network to obtain a first briefing vector set and a first category label vector set;
s22, combining the first briefing vector set and the first class label vector set to obtain a training vector set;
in a preferred embodiment of the present invention, the vector transformation network adopts a Bert deep learning network, and after vectorization operation is performed by the Bert deep learning network, the first presentation vector set { x } is obtained1，x2，x3，...，xnAnd the first category label vector set { y }1，y2，y3，...，ykAnd for the first presentation vector set { x }1，x2，x3，...，xnAnd the first category label vector set { y }1，y2，y3，...，ykPerforming a combining operation to obtain the training vector set s (k) { (x)1，y1)，(x2，y2)，...，(xn，yk)}。
To explain further, in the training vector set s (k) { (x)1，y1)，(x2，y2)，...，(xn，yk) In the method, the presentation vector and the category label vector are in one-to-many or one-to-one relationship, so each presentation vector xnMay correspond to a plurality of class label vectors, so the training vectors in the training vector set may also be
In the form of (1).
S23, calculating a probability distribution value set of the training vector set by using a probability distribution function in the original label classification network;
further, the S23 includes: respectively constructing a first vector set and a second vector set according to the training vector set; and calculating a class prototype set of the first vector set, and solving the probability distribution value sets of the class prototype set and the second vector set.
In detail, the constructing and obtaining a first vector set and a second vector set according to the training vector set respectively includes: and selecting training vectors from the training vector set according to a first preset number to obtain the first vector set, and selecting training vectors of a second preset number from the training vector set to obtain the second vector set.
In detail, the first preset number and the second preset number are generally empirical values and are not greater than the data amount of the training vector set.
In a preferred embodiment of the present invention, the class prototype refers to training vectors belonging to the same class label number, and the average value in the training vector set is calculated as follows:
wherein, ckClass prototype representing class label k, S (k) being said set of training vectors, fθ(xi) Representing the ith training vector in said first set of vectors, a vector representation function in said vector transformation network, xiRepresenting the ith training vector, y, in the first set of vectorsiAnd representing a category label vector corresponding to the ith training vector in the first vector set.
In a preferred embodiment of the present invention, the probability distribution value sets of the class prototype set and the second vector set are solved by using the following probability distribution functions:
wherein x isjRepresenting the jth training vector, p, in said second set of vectorsθ(y＝k|xj) A probability distribution value, f, representing that the jth training vector in the second vector set belongs to a class label kθ(xj) Representing the jth training vector in said second vector set, a vector representation function in said vector transformation network, ckA class prototype representing a class label k.
And S24, calculating a minimized target value of the probability distribution value set, adjusting internal parameters of the original label classification network when the minimized target value is larger than a preset threshold value, and returning to S21 until the minimized target value is smaller than or equal to the preset threshold value, so as to obtain the standard label classification network.
In the preferred embodiment of the present invention, a gradient descent algorithm is used to calculate the minimum target value of the probability distribution value set until a preset threshold requirement is met, so as to obtain the standard label classification network.
S3, performing label classification operation on the second brief report set by using the standard label classification network to obtain a second class label set, summarizing the first class label set and the second class label set to obtain a class label set corresponding to the brief report set.
In a preferred embodiment of the present invention, a standard label classification network is used to perform vectorization operations on the second briefing set, and the probability distribution function is used to solve the class label corresponding to each briefing, so as to obtain the second class label set by summarization.
As can be seen from the above, in the preferred embodiment of the present invention, the first presentation set accounts for 40% of the presentation set, the second presentation set accounts for 60% of the presentation set, and the first category tag set corresponding to the first presentation set and the second category tag set corresponding to the second presentation set are summarized, so as to obtain the category tag set corresponding to the presentation set.
And S4, dividing the briefing set into different categories by using the category label set to obtain a category briefing set, and generating a briefing template by using the category briefing set.
As stated in S2, the category labels correspond to the presentations in a many-to-one or one-to-one correspondence relationship, and if the first sentence of the presentation a is the category 1, and the second sentence and the third sentence are the category 2, the category label set and the presentation set have a many-to-one or one-to-one correspondence relationship, so that each word or the corresponding category label of each segment in each presentation in the presentation set is divided, and the category presentation set is obtained by summarizing.
Further, referring to fig. 3, the generating a presentation template by using the category presentation set includes:
s41, performing attribute replacement operation and part-of-speech tagging operation on the category brief report set to obtain a primary brief report set;
and S42, carrying out sentence sequencing on the primary briefing set to obtain the briefing template.
In detail, the attribute replacement operation adopts a multi-mode trie tree matching method, for example, specific names appearing in the category brief report set are uniformly replaced by 'name' attributes, telephone numbers in a digital form are uniformly replaced by 'telephone', addresses are uniformly replaced by 'cities', for example, 'three persons in innovation centers have cough and running nose symptoms, the current isolation at home is that three persons in the city are 17731897168', and after the attribute replacement operation is executed, the following steps are performed: the < department > employee < name > has symptoms of fever and running nose, and the < name > telephone is a < telephone number > isolated at home at present.
The part-of-speech tagging operation adopts a natural language processing method such as HanLP, wherein w represents punctuation marks, n is a noun, v is a verb, and t is a time word, for example, a word expression form (a < department > employee < name > has symptoms of fever and running nose, and is isolated at home) is converted into a primary brief including (wnwnwnwvvwnnwtvv) part-of-speech arrangement.
The sequence of the sentences is also important for readability of the presentation, and in the preferred embodiment of the invention, the WER value of each primary presentation is calculated by using a WER method (Word Error Rate), and the presentation template is obtained by sequencing the sentences of the primary presentation set according to the WER value of each primary presentation.
S5, receiving the to-be-generated brief report text, executing named entity recognition on the to-be-generated brief report text to obtain an entity set, and executing combined operation on the entity set and the brief report template to obtain the brief report.
In a preferred embodiment of the present invention, the named entity recognition may adopt a bidirectional cyclic neural network BiLSTM, a conditional random field CRF model, etc., and further, according to a specific scenario of the brief report, a corresponding entity set is obtained through the named entity recognition, for example, a specific scenario of the novel coronavirus pneumonia, and the entity set includes a name, an organization, a department, a telephone, an address, a body temperature, symptoms, etc.
Further, according to the part-of-speech arrangement and keyword information corresponding to the entity set, a combination operation is executed with the briefing template, such that the entity with the part-of-speech arrangement (wnwnwvvwnwtvv) exists in the entity set
And the part-of-speech arrangement of the template is the same as that of a certain template of the presentation template, and the keywords in the presentation set corresponding to the presentation template have a matching relationship with the keywords of the presentation text to be generated, the entity sets are sequentially filled into the presentation template to obtain the presentation corresponding to the presentation text to be generated.
The embodiment of the invention divides the brief report set into two to obtain a first brief report set and a second brief report set, trains a label classification network by utilizing the first brief report set and the corresponding first class label set, performs label classification on the second brief report set by utilizing the trained label classification network to obtain a second class label set, and summarizes the first class label set and the second class label set to obtain the class label set corresponding to the brief report set. Because the embodiment of the invention only uses part of the brief report set training model and classifies the rest brief report sets based on the trained model, compared with relying on the model training using all the brief report training data, the embodiment of the invention not only reduces the label of the brief report training data in the early stage, but also reduces the model calculation amount and saves the computer resources because of the reduction of the training data. Therefore, the method, the device and the computer readable storage medium for generating the text briefing can save computing resources and improve the accuracy of briefing generation.
Fig. 4 is a schematic block diagram of a text presentation generating apparatus according to the present invention.
The apparatus 100 for generating a text presentation according to the present invention may be installed in an electronic device. According to the realized functions, the text briefing generating device can comprise a briefing dividing module 101, a classification network training module 102, a class label identification module 103 and a briefing generating module 104. A module according to the present invention, which may also be referred to as a unit, refers to a series of computer program segments that can be executed by a processor of an electronic device and that can perform a fixed function, and that are stored in a memory of the electronic device.
In the present embodiment, the functions regarding the respective modules/units are as follows:
the briefing dividing module 101 is configured to receive a briefing set, classify the briefing set according to a preset proportion, and obtain a first briefing set and a second briefing set;
the classification network training module 102 is configured to receive a first class label set corresponding to the first brief report set, train a pre-constructed original label classification network by using the first brief report set and the first class label set, and obtain a standard label classification network;
the category label identification module 103 is configured to perform a label classification operation on the second presentation set by using the standard label classification network to obtain a second category label set, and summarize the first category label set and the second category label set to obtain a category label set corresponding to the presentation set;
the briefing generating module 104 is configured to divide the briefing set into different categories by using the category tag set to obtain a category briefing set, generate a briefing template by using the category briefing set, receive a briefing text to be generated, perform named entity identification on the briefing text to be generated to obtain an entity set, and perform a combining operation on the entity set and the briefing template to obtain a briefing.
In detail, the specific implementation of each module of the driving violation detection device is as follows:
the briefing dividing module 101 is configured to receive a briefing set, classify the briefing set according to a preset proportion, and obtain a first briefing set and a second briefing set.
The briefing set is a set of briefing under a specific scene, wherein the specific scene comprises a health field briefing scene, an enterprise analysis briefing scene and the like. Taking a health field briefing scene as an example, under the current broad background of the prevalence of the novel coronavirus pneumonia, how to efficiently and accurately extract brief information from massive and various articles to form a briefing for a user to refer is very important, so that the embodiment of the invention obtains a briefing set related to the novel coronavirus pneumonia from a website, a database and the like by using a crawler or a user collection and arrangement mode and the like.
The preset ratio is generally specified by a user, and in a preferred embodiment of the present invention, the preset ratio is 4:6, wherein the data amount of the first presentation set accounts for 40% of the presentation set, and the data amount of the second presentation set accounts for 60% of the presentation set.
The classification network training module 102 is configured to receive a first class label set corresponding to the first presentation set, train a pre-constructed original label classification network by using the first presentation set and the first class label set, and obtain a standard label classification network.
Preferably, the category labels in the first category label set are determined according to specific scenes, and for a new coronary pneumonia brief report scene, the category labels include category 1 "suspected or confirmed by oneself", category 2 "whether suspected or confirmed by oneself is touched, category 3" fever and discomfort symptom by oneself ", and the like. The first category tag set is generally determined by an expert in a specific scene, and corresponds to the first brief report set in a many-to-one or one-to-one manner, for example, brief reports a exist in the first brief report set, the first sentence of the brief report a is category 1, and the second sentence and the third sentence are category 2.
Further, the original label classification network comprises a vector transformation network and a probability distribution function. In detail, the training of the pre-constructed original label classification network by using the first briefing set and the first class label set to obtain a standard label classification network includes:
step A: utilizing a vector conversion network in the original label classification network to perform vectorization operation on the first briefing set and the first category label set to obtain a first briefing vector set and a first category label vector set;
and B: combining the first briefing vector set and the first class label vector set to obtain a training vector set;
and C: calculating a probability distribution value set of the training vector set by using a probability distribution function in the original label classification network;
step D: and B, calculating a minimized target value of the probability distribution value set, adjusting internal parameters of the original label classification network when the minimized target value is larger than a preset threshold value, and returning to the step A until the minimized target value is smaller than or equal to the preset threshold value to obtain the standard label classification network.
In a preferred embodiment of the present invention, the vector transformation network adopts a Bert deep learning network, and after vectorization operation is performed by the Bert deep learning network, the first presentation vector set { x } is obtained1，x2，x3，...，xnAnd the first category label vector set { y }1，y2，y3，...，ykAnd for the first presentation vector set { x }1，x2，x3，...，xnAnd the first category label vector set { y }1，y2，y3，...，ykPerforming a combining operation to obtain the training vector set s (k) { (x)1，y1)，(x2，y2)，...，(xn，yk)}。
To explain further, in the training vector set s (k) { (x)1，y1)，(x2，y2)，...，(xn，yk) In the method, the presentation vector and the category label vector are in one-to-many or one-to-one relationship, so each presentation vector xnMay correspond to a plurality of class label vectors, so the training vectors in the training vector set may also be
In the form of (1).
Further, the calculating a set of probability distribution values of the training vector set using the probability distribution function in the original label classification network includes: respectively constructing a first vector set and a second vector set according to the training vector set; and calculating a class prototype set of the first vector set, and solving the probability distribution value sets of the class prototype set and the second vector set.
In detail, the constructing and obtaining a first vector set and a second vector set according to the training vector set respectively includes: and selecting training vectors from the training vector set according to a first preset number to obtain the first vector set, and selecting training vectors of a second preset number from the training vector set to obtain the second vector set.
In detail, the first preset number and the second preset number are generally empirical values and are not greater than the data amount of the training vector set.
In a preferred embodiment of the present invention, the class prototype refers to training vectors belonging to the same class label number, and the average value in the training vector set is calculated as follows:
wherein, ckClass prototype representing class label k, S (k) being said set of training vectors, fθ(xi) Representing the ith training vector in said first set of vectors, a vector representation function in said vector transformation network, xiRepresenting the ith training vector, y, in the first set of vectorsiAnd representing a category label vector corresponding to the ith training vector in the first vector set.
In a preferred embodiment of the present invention, the probability distribution value sets of the class prototype set and the second vector set are solved by using the following probability distribution functions:
wherein x isjRepresenting the jth training vector, p, in said second set of vectorsθ(y＝k|xj) A probability distribution value, f, representing that the jth training vector in the second vector set belongs to a class label kθ(xj) Representing the jth training vector in said second vector set, a vector representation function in said vector transformation network, ckA class prototype representing a class label k.
In the preferred embodiment of the present invention, a gradient descent algorithm is used to calculate the minimum target value of the probability distribution value set until a preset threshold requirement is met, so as to obtain the standard label classification network.
The category label identification module 103 is configured to perform a label classification operation on the second presentation set by using the standard label classification network to obtain a second category label set, and summarize the first category label set and the second category label set to obtain a category label set corresponding to the presentation set.
In a preferred embodiment of the present invention, a standard label classification network is used to perform vectorization operations on the second briefing set, and the probability distribution function is used to solve the class label corresponding to each briefing, so as to obtain the second class label set by summarization.
As can be seen from the above, in the preferred embodiment of the present invention, the first presentation set accounts for 40% of the presentation set, the second presentation set accounts for 60% of the presentation set, and the first category tag set corresponding to the first presentation set and the second category tag set corresponding to the second presentation set are summarized, so as to obtain the category tag set corresponding to the presentation set.
The briefing generating module 104 is configured to divide the briefing set into different categories by using the category tag set to obtain a category briefing set, generate a briefing template by using the category briefing set, receive a briefing text to be generated, perform named entity identification on the briefing text to be generated to obtain an entity set, and perform a combining operation on the entity set and the briefing template to obtain a briefing.
The category labels and the briefs are in a many-to-one or one-to-one corresponding relationship, if the first sentence of the briefs A is the category 1, and the second sentence and the third sentence are the category 2, each sentence or each section of the corresponding category label in each briefs in the briefs set is divided through the many-to-one or one-to-one corresponding relationship between the category label set and the briefs set, and the category briefs set is obtained through summarization.
Further, the generating a presentation template by using the category presentation set includes: performing attribute replacement operation and part-of-speech tagging operation on the category briefing set to obtain a primary briefing set; and sequencing the sentences of the primary briefing set to obtain the briefing template.
In detail, the attribute replacement operation adopts a multi-mode trie tree matching method, for example, specific names appearing in the category brief report set are uniformly replaced by 'name' attributes, telephone numbers in a digital form are uniformly replaced by 'telephone', addresses are uniformly replaced by 'cities', for example, 'three persons in innovation centers have cough and running nose symptoms, the current isolation at home is that three persons in the city are 17731897168', and after the attribute replacement operation is executed, the following steps are performed: the < department > employee < name > has symptoms of fever and running nose, and the < name > telephone is a < telephone number > isolated at home at present.
The part-of-speech tagging operation adopts a natural language processing method such as HanLP, wherein w represents punctuation marks, n is a noun, v is a verb, and t is a time word, for example, a word expression form (a < department > employee < name > has symptoms of fever and running nose, and is isolated at home) is converted into a primary brief including (wnwnwnwvvwnnwtvv) part-of-speech arrangement.
The sequence of the sentences is also important for readability of the presentation, and in the preferred embodiment of the invention, the WER value of each primary presentation is calculated by using a WER method (Word Error Rate), and the presentation template is obtained by sequencing the sentences of the primary presentation set according to the WER value of each primary presentation.
In a preferred embodiment of the present invention, the named entity recognition may adopt a bidirectional cyclic neural network BiLSTM, a conditional random field CRF model, etc., and further, according to a specific scenario of the brief report, a corresponding entity set is obtained through the named entity recognition, for example, a specific scenario of the novel coronavirus pneumonia, and the entity set includes a name, an organization, a department, a telephone, an address, a body temperature, symptoms, etc.
Further, according to the part-of-speech arrangement and keyword information corresponding to the entity set, a combination operation is executed with the briefing template, such that the entity with the part-of-speech arrangement (wnwnwvvwnwtvv) exists in the entity set
And the part-of-speech arrangement of the template is the same as that of a certain template of the presentation template, and the keywords in the presentation set corresponding to the presentation template have a matching relationship with the keywords of the presentation text to be generated, the entity sets are sequentially filled into the presentation template to obtain the presentation corresponding to the presentation text to be generated.
Fig. 5 is a schematic structural diagram of an electronic device implementing the method for generating a text presentation according to the present invention.
The electronic device 1 may comprise a processor 10, a memory 11 and a bus, and may further comprise a computer program, such as a text briefing generating program 12, stored in the memory 11 and executable on the processor 10.
The memory 11 includes at least one type of readable storage medium, which includes flash memory, removable hard disk, multimedia card, card-type memory (e.g., SD or DX memory, etc.), magnetic memory, magnetic disk, optical disk, etc. The memory 11 may in some embodiments be an internal storage unit of the electronic device 1, such as a removable hard disk of the electronic device 1. The memory 11 may also be an external storage device of the electronic device 1 in other embodiments, such as a plug-in mobile hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like, which are provided on the electronic device 1. Further, the memory 11 may also include both an internal storage unit and an external storage device of the electronic device 1. The memory 11 may be used not only to store application software installed in the electronic device 1 and various types of data, such as codes of the text briefing generating program 12, etc., but also to temporarily store data that has been output or is to be output.
The processor 10 may be composed of an integrated circuit in some embodiments, for example, a single packaged integrated circuit, or may be composed of a plurality of integrated circuits packaged with the same or different functions, including one or more Central Processing Units (CPUs), microprocessors, digital Processing chips, graphics processors, and combinations of various control chips. The processor 10 is a Control Unit (Control Unit) of the electronic device, connects various components of the electronic device by using various interfaces and lines, and executes various functions and processes data of the electronic device 1 by running or executing programs or modules (for example, executing a text presentation generation program, etc.) stored in the memory 11 and calling data stored in the memory 11.
The bus may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is arranged to enable connection communication between the memory 11 and at least one processor 10 or the like.
Fig. 5 only shows an electronic device with components, and it will be understood by a person skilled in the art that the structure shown in fig. 5 does not constitute a limitation of the electronic device 1, and may comprise fewer or more components than shown, or a combination of certain components, or a different arrangement of components.
For example, although not shown, the electronic device 1 may further include a power supply (such as a battery) for supplying power to each component, and preferably, the power supply may be logically connected to the at least one processor 10 through a power management device, so as to implement functions of charge management, discharge management, power consumption management, and the like through the power management device. The power supply may also include any component of one or more dc or ac power sources, recharging devices, power failure detection circuitry, power converters or inverters, power status indicators, and the like. The electronic device 1 may further include various sensors, a bluetooth module, a Wi-Fi module, and the like, which are not described herein again.
Further, the electronic device 1 may further include a network interface, and optionally, the network interface may include a wired interface and/or a wireless interface (such as a WI-FI interface, a bluetooth interface, etc.), which are generally used for establishing a communication connection between the electronic device 1 and other electronic devices.
Optionally, the electronic device 1 may further comprise a user interface, which may be a Display (Display), an input unit (such as a Keyboard), and optionally a standard wired interface, a wireless interface. Alternatively, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, or the like. The display, which may also be referred to as a display screen or display unit, is suitable for displaying information processed in the electronic device 1 and for displaying a visualized user interface, among other things.
It is to be understood that the described embodiments are for purposes of illustration only and that the scope of the appended claims is not limited to such structures.
The text briefing generating program 12 stored in the memory 11 of the electronic device 1 is a combination of instructions that, when executed in the processor 10, enable:
receiving a brief report set, and classifying the brief report set according to a preset proportion to obtain a first brief report set and a second brief report set;
receiving a first class label set corresponding to the first brief report set, and training a pre-constructed original label classification network by using the first brief report set and the first class label set to obtain a standard label classification network;
performing label classification operation on the second brief report set by using the standard label classification network to obtain a second class label set, summarizing the first class label set and the second class label set to obtain a class label set corresponding to the brief report set;
dividing the briefing set into different categories by using the category label set to obtain a category briefing set, and generating a briefing template by using the category briefing set;
receiving a to-be-generated brief report text, executing named entity recognition on the to-be-generated brief report text to obtain an entity set, and executing combined operation on the entity set and the brief report template to obtain a brief report.
Further, the integrated modules/units of the electronic device 1, if implemented in the form of software functional units and sold or used as separate products, may be stored in a computer readable storage medium. The computer-readable medium may include: any entity or device capable of carrying said computer program code, recording medium, U-disk, removable hard disk, magnetic disk, optical disk, computer Memory, Read-Only Memory (ROM).
Further, the computer usable storage medium may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function, and the like; the storage data area may store data created according to the use of the blockchain node, and the like.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus, device and method can be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is only one logical functional division, and other divisions may be realized in practice.
The modules described as separate parts may or may not be physically separate, and parts displayed as modules may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment.
In addition, functional modules in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional module.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof.
The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any accompanying claims should not be construed as limiting the claim concerned.
The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.
Furthermore, it is obvious that the word "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. A plurality of units or means recited in the system claims may also be implemented by one unit or means in software or hardware. The terms second, etc. are used to denote names, but not any particular order.
Finally, it should be noted that the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting, and although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims (10)
1. A method for generating a text presentation, the method comprising:
receiving a brief report set, and classifying the brief report set according to a preset proportion to obtain a first brief report set and a second brief report set;
receiving a first class label set corresponding to the first brief report set, and training a pre-constructed original label classification network by using the first brief report set and the first class label set to obtain a standard label classification network;
performing label classification operation on the second brief report set by using the standard label classification network to obtain a second class label set, summarizing the first class label set and the second class label set to obtain a class label set corresponding to the brief report set;
dividing the briefing set into different categories by using the category label set to obtain a category briefing set, and generating a briefing template by using the category briefing set;
receiving a to-be-generated brief report text, executing named entity recognition on the to-be-generated brief report text to obtain an entity set, and executing combined operation on the entity set and the brief report template to obtain a brief report.
2. The method of claim 1, wherein the training of a pre-constructed original tag classification network using the first presentation set and the first category tag set to obtain a standard tag classification network comprises:
respectively executing vectorization operation on the first briefing set and the first category label set by using a vector conversion network in the original label classification network to obtain a first briefing vector set and a first category label vector set;
performing a combination operation on the first briefing vector set and the first class label vector set to obtain a training vector set;
calculating a probability distribution value set of the training vector set by using a probability distribution function in the original label classification network;
and calculating a minimized target value of the probability distribution value set, and when the minimized target value is greater than a preset threshold value, adjusting internal parameters of the original label classification network until the minimized target value is less than or equal to the preset threshold value, so as to obtain the standard label classification network.
3. The method of generating a text presentation of claim 2 wherein the computing a set of probability distribution values for the set of training vectors using the probability distribution functions in the original label classification network comprises:
respectively constructing a first vector set and a second vector set according to the training vector set;
calculating a set of class prototypes for the first set of vectors;
and solving to obtain the probability distribution value set by taking the class prototype set and the second vector set as function parameters of the probability distribution function.
4. The method of generating a text presentation of claim 3, wherein said computing a set of class prototypes for the first set of vectors comprises:
calculating a class prototype set of the first vector set by adopting the following method:
wherein, ckClass prototypes representing class labels k, S (k) being the training directionVolume set, fθ(xi) A vector representation function, x, representing the ith training vector in the first vector set in the vector transformation networkiRepresenting the ith training vector, y, in the first set of vectorsiAnd representing a category label vector corresponding to the ith training vector in the first vector set.
5. The method of generating a text presentation according to claim 4, wherein said solving a set of probability distribution values for said class prototype set and said second vector set using said probability distribution function comprises:
solving a set of probability distribution values of the class prototype set and the second vector set by using the following probability distribution functions:
wherein x isjRepresenting the jth training vector, p, in said second set of vectorsθ(y＝k|xj) A probability distribution value, f, representing that the jth training vector in the second vector set belongs to a class label kθ(xj) Representing a vector representation function of a jth training vector of said second set of vectors in said vector transformation network, ckA class prototype representing a class label k.
6. The method of generating a text presentation as recited in claim 1, wherein the generating a presentation template using the set of category presentations comprises:
performing attribute replacement operation and part-of-speech tagging operation on the category briefing set to obtain a primary briefing set;
and sequencing the sentences of the primary briefing set to obtain the briefing template.
7. The method for generating a text presentation according to any one of claims 1 to 6, wherein the preset ratio is 4: 6.
8. An apparatus for generating a text presentation, the apparatus comprising:
the briefing dividing module is used for receiving a briefing set, classifying the briefing set according to a preset proportion, and obtaining a first briefing set and a second briefing set;
the classification network training module is used for receiving a first class label set corresponding to the first brief report set, and training a pre-constructed original label classification network by using the first brief report set and the first class label set to obtain a standard label classification network;
a category label identification module, configured to perform a label classification operation on the second brief report set by using the standard label classification network to obtain a second category label set, and summarize the first category label set and the second category label set to obtain a category label set corresponding to the brief report set;
the briefing generating module is used for dividing the briefing set into different categories by using the category label set to obtain a category briefing set, generating a briefing template by using the category briefing set, receiving a briefing text to be generated, performing named entity identification on the briefing text to be generated to obtain an entity set, and performing combined operation on the entity set and the briefing template to obtain the briefing.
9. An electronic device, characterized in that the electronic device comprises:
at least one processor; and the number of the first and second groups,
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 a method of generating a text presentation according to any one of claims 1 to 7.
10. A computer-readable storage medium comprising a storage data area and a storage program area, wherein the storage data area stores created data, and the storage program area stores a computer program; wherein the computer program, when executed by a processor, implements a method of generating a text briefing as claimed in any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011054465.7A CN112214594A (en) | 2020-09-30 | 2020-09-30 | Text briefing generation method and device, electronic equipment and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011054465.7A CN112214594A (en) | 2020-09-30 | 2020-09-30 | Text briefing generation method and device, electronic equipment and readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112214594A true CN112214594A (en) | 2021-01-12 |
Family
ID=74052156
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011054465.7A Withdrawn CN112214594A (en) | 2020-09-30 | 2020-09-30 | Text briefing generation method and device, electronic equipment and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112214594A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114022709A (en) * | 2021-11-04 | 2022-02-08 | 内蒙古呼和浩特市立信电气技术有限责任公司 | Control method and system based on stockpile layered three-dimensional real-time model |
-
2020
- 2020-09-30 CN CN202011054465.7A patent/CN112214594A/en not_active Withdrawn
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114022709A (en) * | 2021-11-04 | 2022-02-08 | 内蒙古呼和浩特市立信电气技术有限责任公司 | Control method and system based on stockpile layered three-dimensional real-time model |
| CN114022709B (en) * | 2021-11-04 | 2024-11-01 | 内蒙古呼和浩特市立信电气技术有限责任公司 | Control method and system based on material pile layering three-dimensional real-time model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111639153A (en) | Query method and device based on legal knowledge graph, electronic equipment and medium | |
| CN113704429A (en) | Semi-supervised learning-based intention identification method, device, equipment and medium | |
| CN112883190A (en) | Text classification method and device, electronic equipment and storage medium | |
| CN115002200A (en) | User portrait based message pushing method, device, equipment and storage medium | |
| CN112559687A (en) | Question identification and query method and device, electronic equipment and storage medium | |
| CN113821622A (en) | Answer retrieval method and device based on artificial intelligence, electronic equipment and medium | |
| CN113378970A (en) | Sentence similarity detection method and device, electronic equipment and storage medium | |
| CN112988963A (en) | User intention prediction method, device, equipment and medium based on multi-process node | |
| CN112885423A (en) | Disease label detection method and device, electronic equipment and storage medium | |
| CN112528013A (en) | Text abstract extraction method and device, electronic equipment and storage medium | |
| CN114491047A (en) | Multi-label text classification method and device, electronic equipment and storage medium | |
| CN113887941A (en) | Business process generation method and device, electronic equipment and medium | |
| CN113807973A (en) | Text error correction method and device, electronic equipment and computer readable storage medium | |
| CN114595321A (en) | Question marking method and device, electronic equipment and storage medium | |
| CN114880449A (en) | Reply generation method and device of intelligent question answering, electronic equipment and storage medium | |
| CN114840684A (en) | Map construction method, device and equipment based on medical entity and storage medium | |
| CN114003704A (en) | Method and device for creating designated tag guest group, electronic equipment and storage medium | |
| CN113434650A (en) | Question and answer pair expansion method and device, electronic equipment and readable storage medium | |
| CN112214594A (en) | Text briefing generation method and device, electronic equipment and readable storage medium | |
| CN112347739A (en) | Application rule analysis method and device, electronic equipment and storage medium | |
| CN112214602A (en) | Text classification method and device based on humor, electronic equipment and storage medium | |
| CN116521867A (en) | Text clustering method and device, electronic equipment and storage medium | |
| CN115146064A (en) | Intention recognition model optimization method, device, equipment and storage medium | |
| CN115221274A (en) | Text emotion classification method and device, electronic equipment and storage medium | |
| CN114385815A (en) | News screening method, device, equipment and storage medium based on business requirements |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20210112 |