CN113111174A - Group identification method, device, equipment and medium based on deep learning model - Google Patents
Group identification method, device, equipment and medium based on deep learning model Download PDFInfo
- Publication number
- CN113111174A CN113111174A CN202010348490.XA CN202010348490A CN113111174A CN 113111174 A CN113111174 A CN 113111174A CN 202010348490 A CN202010348490 A CN 202010348490A CN 113111174 A CN113111174 A CN 113111174A
- Authority
- CN
- China
- Prior art keywords
- group
- word
- descriptor
- text
- descriptors
- 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
- 238000013136 deep learning model Methods 0.000 title claims abstract description 62
- 238000000034 method Methods 0.000 title claims abstract description 54
- 230000011218 segmentation Effects 0.000 claims abstract description 103
- 238000012549 training Methods 0.000 claims abstract description 91
- 239000013598 vector Substances 0.000 claims abstract description 78
- 238000012545 processing Methods 0.000 claims abstract description 61
- 238000013145 classification model Methods 0.000 claims abstract description 48
- 238000002372 labelling Methods 0.000 claims description 23
- 238000004590 computer program Methods 0.000 claims description 10
- 238000000605 extraction Methods 0.000 abstract description 4
- 230000008569 process Effects 0.000 description 11
- 239000003795 chemical substances by application Substances 0.000 description 9
- 238000010586 diagram Methods 0.000 description 9
- 238000004891 communication Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 7
- 230000008520 organization Effects 0.000 description 6
- 238000013527 convolutional neural network Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 238000011176 pooling Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 239000000835 fiber Substances 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000036632 reaction speed Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000004065 semiconductor Substances 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
- G06F16/35—Clustering; Classification
-
- 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
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
-
- 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
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/18—Legal services
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Tourism & Hospitality (AREA)
- Computing Systems (AREA)
- Technology Law (AREA)
- Databases & Information Systems (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The disclosure provides a group identification method and device based on a deep learning model, equipment and a medium. One embodiment of the method comprises: performing word segmentation on the alarm receiving and processing text to be recognized to obtain a corresponding word segmentation sequence; for each participle in the obtained participle sequence, inputting a word vector corresponding to the participle into a group descriptor classification model to determine whether the participle is a group descriptor or not, wherein the group descriptor classification model is obtained by pre-training based on a deep learning model, and the group descriptor is a word in a group description text for describing a group; generating a group description text by using a segmentation sequence segment consisting of continuous adjacent group description words in the obtained segmentation sequence; and generating a group description text set corresponding to the alarm receiving and processing text to be identified by using the generated group description texts. The embodiment realizes the automatic extraction of the group description text in the alarm receiving and processing text.
Description
Technical Field
The present disclosure relates to the field of computer technologies, and in particular, to a group recognition method and apparatus, a device, and a medium based on a deep learning model.
Background
The police department can generate an alarm receiving text after receiving an alarm and can generate an alarm handling text after handling the alarm. The alarm receiving and processing text comprises the alarm receiving text and the alarm processing text. In practice, some of the alarm-receiving texts may involve descriptions about groups or group events. Here, the population refers to a population that spontaneously gathers together for the same or similar purposes. There is no strict organizational management system in the population. Such as college entrance students, owner groups of a certain cell, etc. For public security, different groups or group events need to be supervised, tracked and investigated in different ways. Therefore, it is important for the public security organization to identify the group through the alarm receiving and processing text, that is, it is important to extract the group description text for describing the group in the alarm receiving and processing text.
However, at present, the group description text in the alarm receiving and processing text is basically extracted manually, and the cost of manpower and time is high. When a group or a group event needing to be paid attention to tracking processing in time occurs, the group or the group event cannot be found in time and follow-up processing in time due to low manual extraction speed. In addition, most of the alarm receiving and processing texts adopt natural language description and have serious and irregular expression modes, so that the manual extraction difficulty is high, and the learning cost of the process of manually extracting the group description texts in the alarm receiving and processing texts is high depending on manual experience.
Disclosure of Invention
The disclosure provides a group identification method and device based on a deep learning model, equipment and a medium.
In a first aspect, the present disclosure provides a deep learning model-based population identification method, including: performing word segmentation on the alarm receiving and processing text to be recognized to obtain a corresponding word segmentation sequence; for each participle in the obtained participle sequence, inputting a word vector corresponding to the participle into a group descriptor classification model to determine whether the participle is a group descriptor or not, wherein the group descriptor classification model is obtained by pre-training based on a deep learning model, and the group descriptor is a word in a group description text for describing a group; generating a group description text by using a segmentation sequence segment consisting of continuous adjacent group description words in the obtained segmentation sequence; and generating a group description text set corresponding to the alarm receiving and processing text to be identified by using the generated group description texts.
In some alternative embodiments, the population descriptor classification model is obtained by pre-training through the following training steps: acquiring a training sample set, wherein the training sample comprises a segmentation sequence obtained by segmenting a historical alarm receiving and processing text and a corresponding labeling information sequence, and the labeling information in the labeling information sequence is used for indicating whether a corresponding segmentation in the corresponding segmentation sequence belongs to a group description text included in the corresponding historical alarm receiving and processing text; determining a group descriptor and a non-group descriptor in the word segmentation sequence of each training sample of the training sample set according to the labeled information sequence in each training sample of the training sample set; generating a positive sample set and a negative sample set, wherein the positive sample comprises a word vector corresponding to the determined population descriptors and an annotation classification result used for indicating that the population descriptors are the population descriptors, and the negative sample comprises a word vector corresponding to the determined non-population descriptors and an annotation classification result used for indicating the non-population descriptors; and training an initial deep learning model by taking the word vectors in the positive sample set and the negative sample set as actual input and taking corresponding labeled classification results as expected output to obtain the group descriptor classification model.
In some optional embodiments, each component in the word vector corresponding to the group descriptor and each component in the word vector corresponding to the non-group descriptor respectively correspond to each word in the preset dictionary in a one-to-one manner, a component in the word vector corresponding to the group descriptor in the group descriptor corresponding to the group descriptor is a word frequency-inverse text frequency index of the group descriptor, a component different from the component corresponding to the group descriptor is a first preset value, a component in the word vector corresponding to the non-group descriptor in the non-group descriptor corresponding to the non-group descriptor is a word frequency-inverse text frequency index of the non-group descriptor, and a component different from the component corresponding to the non-group descriptor is the first preset value.
In some alternative embodiments, a ratio of the number of positive samples in the positive sample set divided by the number of negative samples in the negative sample set is within a preset ratio.
In a second aspect, the present disclosure provides a deep learning model-based group recognition apparatus, including: the word segmentation unit is configured to segment words of the alarm receiving and processing text to be identified to obtain a corresponding word segmentation sequence; a classification unit configured to, for each participle in the obtained participle sequence, input a word vector corresponding to the participle into a group descriptor classification model to determine whether the participle is a group descriptor, wherein the group descriptor classification model is obtained by pre-training based on a deep learning model, and the group descriptor is a word in a group description text for describing a group; a first generating unit configured to generate a group description text using a segmentation sequence segment composed of consecutive adjacent group description words in the obtained segmentation sequence; and the second generating unit is configured to generate a group description text set corresponding to the alarm receiving and processing text to be identified by using each generated group description text.
In some alternative embodiments, the population descriptor classification model is obtained by pre-training through the following training steps: acquiring a training sample set, wherein the training sample comprises a word segmentation sequence obtained by segmenting a historical alarm receiving text and a corresponding labeling information sequence, and the labeling information in the labeling information sequence is used for indicating whether a corresponding word in the corresponding word segmentation sequence is a group descriptor in a group description text included in the corresponding historical alarm receiving text; determining a group descriptor and a non-group descriptor in the word segmentation sequence of each training sample of the training sample set according to the labeled information sequence in each training sample of the training sample set; generating a positive sample set and a negative sample set, wherein the positive sample comprises a word vector corresponding to the determined population descriptors and an annotation classification result used for indicating that the population descriptors are the population descriptors, and the negative sample comprises a word vector corresponding to the determined non-population descriptors and an annotation classification result used for indicating the non-population descriptors; and training an initial deep learning model by taking the word vectors in the positive sample set and the negative sample set as actual input and taking corresponding labeled classification results as expected output to obtain the group descriptor classification model.
In some optional embodiments, each component in the word vector corresponding to the group descriptor and each component in the word vector corresponding to the non-group descriptor respectively correspond to each word in the preset dictionary in a one-to-one manner, a component in the word vector corresponding to the group descriptor in the group descriptor corresponding to the group descriptor is a word frequency-inverse text frequency index of the group descriptor, a component different from the component corresponding to the group descriptor is a first preset value, a component in the word vector corresponding to the non-group descriptor in the non-group descriptor corresponding to the non-group descriptor is a word frequency-inverse text frequency index of the non-group descriptor, and a component different from the component corresponding to the non-group descriptor is the first preset value.
In some alternative embodiments, a ratio of the number of positive samples in the positive sample set divided by the number of negative samples in the negative sample set is within a preset ratio.
In a third aspect, the present disclosure provides an electronic device, comprising: one or more processors; a storage device, on which one or more programs are stored, which, when executed by the one or more processors, cause the one or more processors to implement the method as described in any implementation manner of the first aspect.
In a fourth aspect, the present disclosure provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by one or more processors, implements the method as described in any of the implementations of the first aspect.
According to the group identification method and device based on the deep learning model, the corresponding word segmentation sequence is obtained by firstly segmenting the alarm receiving and processing text to be identified. And inputting a word vector corresponding to the participle into the group descriptor classification model for each participle in the obtained participle sequence so as to determine whether the participle is a group descriptor. Then, a group description text is generated by using a segmentation word sequence segment composed of continuous adjacent group description words in the obtained segmentation word sequence. And finally, generating a group description text set corresponding to the alarm receiving and processing text to be identified by using the generated group description texts. The whole process does not need manual operation, and labor cost and time cost for generating the group description text set corresponding to the alarm receiving and processing text to be identified are reduced. When the group identification method and the group identification device are applied to processing a large number of recently (for example, within one day) generated alarm receiving texts, the group description text set can be quickly extracted from the large number of alarm receiving texts. With the extracted group description texts, the public security organization can perform corresponding processing according to the obtained group description texts in time, so that the processing response speed of the public security organization to the groups and the group events is improved, and the public security organization can accurately correspond to the groups and the group events.
Drawings
Other features, objects and advantages of the disclosure will become more apparent upon reading of the following detailed description of non-limiting embodiments thereof, made with reference to the accompanying drawings in which:
FIG. 1 is an exemplary system architecture diagram in which one embodiment of the present disclosure may be applied;
FIG. 2 is a flow diagram of one embodiment of a deep learning model-based population identification method according to the present disclosure;
FIG. 3 is a flow chart of one embodiment of training steps according to the present disclosure;
FIG. 4 is a schematic diagram of an embodiment of a deep learning model-based population recognition device according to the present disclosure;
FIG. 5 is a schematic block diagram of a computer system suitable for use in implementing the electronic device of the present disclosure.
Detailed Description
The present disclosure is described in further detail below with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the relevant invention and not restrictive of the invention. It should be noted that, for convenience of description, only the portions related to the related invention are shown in the drawings.
It should be noted that, in the present disclosure, the embodiments and features of the embodiments may be combined with each other without conflict. The present disclosure will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.
Fig. 1 illustrates an exemplary system architecture 100 to which embodiments of the deep learning model-based population recognition method or the deep learning model-based population recognition apparatus of the present disclosure may be applied.
As shown in fig. 1, system architecture 100 may include terminal device 101, network 102, and server 103. Network 102 is the medium used to provide communication links between terminal devices 101 and server 103. Network 102 may include various connection types, such as wired, wireless communication links, or fiber optic cables, to name a few.
A user may use terminal device 101 to interact with server 103 over network 102 to receive or send messages and the like. Various communication client applications, such as an alarm receiving and processing record application, an alarm receiving and processing text group recognition application, a web browser application, etc., may be installed on the terminal device 101.
The terminal apparatus 101 may be hardware or software. When the terminal device 101 is hardware, it may be various electronic devices having a display screen and supporting text input, including but not limited to smart phones, tablet computers, laptop portable computers, desktop computers, and the like. When the terminal apparatus 101 is software, it can be installed in the electronic apparatuses listed above. It may be implemented as multiple software or software modules (e.g., to provide alarm text group recognition services) or as a single software or software module. And is not particularly limited herein.
The server 103 may be a server providing various services, such as a background server providing group identification services for the alarm receiving texts sent by the terminal device 101. The background server can analyze and process the received alarm receiving and processing text, and feed back the processing result (such as the group description text set) to the terminal device.
In some cases, the group recognition method based on the deep learning model provided by the present disclosure may be performed by the terminal device 101 and the server 103 together, for example, the step of "performing word segmentation on the alarm receiving and processing text to be recognized to obtain the corresponding word segmentation sequence" may be performed by the terminal device 101, and the rest of the steps may be performed by the server 103. The present disclosure is not limited thereto. Accordingly, group recognition means based on the deep learning model may also be provided in the terminal device 101 and the server 103, respectively.
In some cases, the group identification method based on deep learning model provided by the present disclosure may be executed by the server 103, and accordingly, the group identification apparatus based on deep learning model may also be disposed in the server 103, and in this case, the system architecture 100 may also not include the terminal device 101.
In some cases, the group identification method based on deep learning model provided by the present disclosure may be executed by the terminal device 101, and accordingly, the group identification apparatus based on deep learning model may also be disposed in the terminal device 101, and in this case, the system architecture 100 may not include the server 103.
The server 103 may be hardware or software. When the server 103 is hardware, it may be implemented as a distributed server cluster composed of a plurality of servers, or may be implemented as a single server. When the server 103 is software, it may be implemented as multiple software or software modules (e.g., to provide alarm handling text crowd identification services) or as a single software or software module. And is not particularly limited herein.
It should be understood that the number of terminal devices, networks, and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.
With continued reference to FIG. 2, a flow 200 of one embodiment of a deep learning model-based population identification method according to the present disclosure is shown. The group identification method based on the deep learning model comprises the following steps:
In the embodiment, an executing subject (for example, a server shown in fig. 1) of the deep learning model-based group recognition method may first obtain an alarm receiving and processing text to be recognized. Then, various word segmentation methods can be adopted to segment the words of the acquired alarm receiving and processing text to be recognized to obtain a corresponding word segmentation sequence.
Here, the execution main body may obtain the locally stored to-be-identified alarm receiving and processing text, or the execution main body may remotely obtain the to-be-identified alarm receiving and processing text from another electronic device (for example, the terminal device shown in fig. 1) connected to the execution main body through a network.
Here, the alarm receiving and processing text to be recognized may be text data that an alarm receiver arranges according to the contents of an alarm receiving telephone or text data that an alarm processor arranges according to an alarm processing procedure. The alarm receiving and processing text to be identified can also be an alarm text which is received from the terminal equipment and is input by a user in an alarm application installed on the terminal equipment or a webpage with an alarm function.
It should be noted that how to cut words of text is the prior art of extensive research and application in this field, and will not be described herein. For example, a word segmentation method based on string matching, a word segmentation method based on understanding, or a word segmentation method based on statistics, etc. may be employed. For example, the word segmentation of the to-be-recognized alarm-receiving text "a lot of first-cell owner complaint about property parking indiscriminate charges" may result in the word segmentation sequence "a lot/first/cell/owner/complaint/property/parking/indiscriminate/charges".
In this embodiment, for each participle in the obtained participle sequence, the executing body may first determine a word vector corresponding to the participle, and then input the word vector corresponding to the participle into the population descriptor classification model to determine whether the participle is a population descriptor. Here, the group descriptor may be a word in a group description text for describing the group.
Here, the population descriptor classification model may be trained in advance based on a deep learning model. The group descriptor classification model is used for representing the corresponding relation between a word vector corresponding to a word and a classification result used for representing whether the word is a group descriptor. It can be understood that, after the word vector corresponding to the participle is input into the group descriptor classification model, if the obtained classification result is used for representing that the participle is a group descriptor, the participle can be determined to be the group descriptor; on the contrary, if a classification result for characterizing that the participle is not a group descriptor is obtained, it can be determined that the participle is not a group descriptor.
As an example, the group descriptor classification model may be obtained by training a deep learning model in advance based on word vectors of a large number of sample words and corresponding labeled sample classification results used for characterizing whether the sample words are group descriptors.
In practice, the execution body may determine the word vector corresponding to the word segmentation by adopting various implementations.
In some optional embodiments, the word vector corresponding to the segmented word may include N-dimensional components, where N is a positive integer, and each of the N-dimensional components corresponds to each word of the preset dictionary one to one. In the process of determining the word vector corresponding to the word segmentation, a component corresponding to the word segmentation in each component of the word vector of the word segmentation may be set to be a second preset value (e.g., 1); the other component of the word vector corresponding to the participle (i.e., the component corresponding to a word in the preset dictionary other than the participle) is set to a third preset numerical value (e.g., 0).
In some optional embodiments, the word vector corresponding to the segmented word may include N-dimensional components, where N is a positive integer, and each of the N-dimensional components corresponds to each word of the preset dictionary one to one. In the process of determining the word vector corresponding to the participle, the execution main body may also first calculate a word Frequency-Inverse text Frequency index (TF-IDF, Term Frequency-Inverse Document Frequency) of the participle in the text to be recognized and processed, set a component corresponding to the participle in the word vector corresponding to the participle as the calculated word Frequency-Inverse text Frequency index of the participle, and finally set other components of the word vector corresponding to the participle (i.e., components corresponding to words different from the participle in the preset dictionary) as a fourth preset numerical value (e.g., 0).
For example, for each participle in a participle sequence "large amount/first/cell/owner/complaint/property/parking/scrambling/charging" corresponding to the alarm receiving text "large amount of first cell owner complaint property parking scrambling charging", a word vector corresponding to the participle is input into a pre-trained group descriptor classification model, and a classification result used for representing whether the participle is a group descriptor is obtained. Referring to table 1, table 1 shows classification results obtained by inputting each participle in the participle sequence into the population descriptor classification model.
TABLE 1
As can be seen from table 1, the sequence of the words "large/first/cell/owner/complaint/property/parking/random/charging" includes four words "large/first/cell/owner" of the group descriptor, but not five words "complaint/property/parking/random/charging" of the group descriptor.
It should be noted that, through step 202, the obtained participle sequence may not include any group descriptors, or the obtained participle sequence may include at least one group descriptor.
And step 203, generating a group description text by using the segmentation sequence segment consisting of the continuous adjacent group description words in the obtained segmentation sequence.
In step 202, it has been determined which words in the sequence of participles obtained in step 201 are population descriptors and which words are not population descriptors, and the participles in the sequence of participles are arranged in the order of their occurrence in the text to be recognized as a receive-and-alarm. If there are consecutive adjacent group descriptors in the segmentation sequence, and the group descriptors are words in the group description text, it can be considered that these consecutive adjacent group descriptors can constitute the group description text. Therefore, in the present embodiment, the execution subject may generate the group description text using a segmentation word sequence segment composed of consecutive adjacent group description words in the obtained segmentation word sequence.
For the sake of understanding, continuing with the above example of the text to be recognized and processed for alarm receiving and classifying results in table 1, the segment of the segmentation sequence consisting of the continuous group descriptors in the segmentation sequence "large/first/cell/owner/complaint/property/parking/random/charging" is "large/first/cell/owner", and the text generated by using the segment of the segmentation sequence "large a-cell owner" is the group description text.
It should be noted that if it is determined that the obtained word segmentation sequence does not include any group descriptor through step 202, the group descriptor text is not generated when step 203 is executed. If the obtained segmentation sequence includes at least one group descriptor after step 202, but the segmentation sequence may include a segmentation sequence segment composed of consecutive adjacent group descriptors, a group description text may be generated by using the included segmentation sequence segment when step 203 is executed. If the word segmentation sequence includes more than one word segmentation sequence segment composed of consecutive adjacent group descriptors, for example, S word segmentation sequence segments, where S is a positive integer greater than or equal to 2, after step 203 is completed, each word segmentation sequence segment in the S word segmentation sequence segments may be used to generate a corresponding group description text, that is, S group description texts may be generated.
And 204, generating a group description text set corresponding to the alarm receiving and processing text to be identified by using the generated group description texts.
In this embodiment, the executing agent may generate a group description text set corresponding to the alarm receiving and processing text to be recognized by using each group description text generated in step 203.
It should be noted that, since the number of the group description texts generated in step 203 may be 0,1 or a positive integer greater than 1, the number of the group description texts that may be included in the group description text set generated in step 204 may also be 0,1 or a positive integer greater than 1. That is, at least one section of group description text describing the group may exist in the alarm receiving and processing text to be recognized, or the group description text describing the group may not exist in the alarm receiving and processing text to be recognized.
In some alternative implementations, the population descriptor classification model may be pre-trained through a training step as shown in fig. 3. Referring to fig. 3, fig. 3 illustrates a flow 300 of one embodiment of training steps according to the present disclosure. The training step comprises the following steps:
here, the execution subject of the training step may be the same as that of the above-described population recognition method based on the deep learning model. In this way, after the population classification model is obtained by training, the executing agent in the training step may store the model parameters of the population classification model in the local executing agent, and read the model structure information and the model parameter information of the population descriptor classification model obtained by training in the process of executing the population recognition method based on the deep learning model.
Here, the execution subject of the training step may be different from that of the group recognition method based on the deep learning model described above. In this way, the executing agent of the training step may send the model parameters of the group descriptor classification model to the executing agent of the group recognition method based on the deep learning model after the group descriptor classification model is obtained through training. In this way, the executing agent of the group recognition method based on the deep learning model may read the model structure information and the model parameter information of the group descriptor classification model received from the executing agent of the training step in the process of executing the group recognition method based on the deep learning model.
Here, the performing subject of the training step may first obtain a set of training samples. Each training sample may include a segmentation sequence obtained by segmenting a historical alarm receiving and processing text and a corresponding tagging information sequence, where tagging information in the tagging information sequence is used to indicate whether a corresponding segmentation in the corresponding segmentation sequence belongs to a group description text included in the corresponding historical alarm receiving and processing text.
As an example, the training sample may include a word segmentation sequence "a lot/a/b/a/s" corresponding to the word segmentation sequence "1/1/1/1/0/0/0/0/0", where "0" is used to indicate that the word corresponding to the position of the word in the word segmentation sequence "a lot/a/s" 1' the participles corresponding to the same positions in the annotation information sequence are population descriptors. That is, among them, the four participles of "mass/first/cell/owner" are group descriptors, and the five participles of "complaint/property/parking/random/charge" are not group descriptors.
In practice, various implementation manners can be adopted to perform word segmentation on the historical alarm receiving and processing text to obtain a word segmentation sequence, and the related records can be referred to, and are not described herein again. And the labeling information sequence corresponding to the word segmentation sequence can be obtained through manual labeling.
Because each training sample comprises a word segmentation sequence and a corresponding tagging information sequence, the tagging information sequence characterizes whether a corresponding word in the corresponding word segmentation sequence belongs to a group description text included in the historical alarm receiving text corresponding to the word segmentation sequence, and if a certain word in the word segmentation sequence belongs to the group description text included in the historical alarm receiving text corresponding to the word segmentation sequence, the word can be considered as a group description word relative to the historical alarm receiving text. If a word in the word segmentation sequence does not belong to the group description text included in the historical alarm receiving text corresponding to the word segmentation sequence, the word can be considered as a non-group description word relative to the historical alarm receiving text. Therefore, the executing body of the training step may determine the group descriptors and the non-group descriptors in the word segmentation sequence of each training sample of the training sample set according to the tagged information sequence in each training sample of the training sample set. It will be appreciated that in practice it may occur that a word is a group descriptor with respect to one historical alarm-receiving text and a non-group descriptor with respect to another historical alarm-receiving text.
Here, the executing agent of the training step may generate the positive sample set and the negative sample set in various ways based on the population descriptors and the non-population descriptors determined in step 302. The positive sample can comprise a word vector corresponding to the determined group descriptor and an annotation classification result used for indicating that the group descriptor is the group descriptor, and the negative sample can comprise a word vector corresponding to the determined non-group descriptor and an annotation classification result used for indicating the non-group descriptor.
In some alternative embodiments, the executing entity of the training step may generate a positive sample set by using the word vectors corresponding to all the group descriptors determined in step 302 and the labeled classification result indicating that the group descriptors are group descriptors, and generate a negative sample set by using the word vectors corresponding to all the non-group descriptors determined in step 302 and the labeled classification result indicating that the non-group descriptors are non-group descriptors.
In some alternative embodiments, in order to control the quantity ratio of the positive samples and the negative samples to achieve a better training effect, the executing entity of the training step may select a corresponding quantity of population descriptors and non-population descriptors from the population descriptors and non-population descriptors determined in step 302 to generate a positive sample set and a negative sample set, and a ratio of the number of positive samples in the generated positive sample set divided by the number of negative samples in the negative sample set is within a preset ratio range. For example, the preset ratio range may be 2 or more and 6 or less. I.e. the number of positive samples is relatively larger than the number of negative samples, but there are also negative samples.
In some optional embodiments, the word vector corresponding to the group descriptor and the word vector corresponding to the non-group descriptor may include N-dimensional components, where N is a positive integer, and each of the N-dimensional components corresponds to each word of the preset dictionary one to one. In the process of determining the word vectors corresponding to the group descriptors and the non-group descriptors, the components corresponding to the group descriptors and the non-group descriptors in the components of the word vectors corresponding to the group descriptors and the non-group descriptors may be set to a second preset value (e.g., 1); the other component of the word vector corresponding to the participle (i.e., the component corresponding to a word in the preset dictionary other than the group descriptor and the non-group descriptor) is set to a third preset numerical value (e.g., 0).
In some optional implementations, each component in the word vector corresponding to the group descriptor and the word vector corresponding to the non-group descriptor may respectively correspond to each word in the preset dictionary in a one-to-one manner, a component in the word vector corresponding to the group descriptor in the group descriptor corresponding to the group descriptor may be a word frequency-inverse text frequency index of the group descriptor, a component different from the component corresponding to the group descriptor may be a first preset value (e.g., 0), a component in the word vector corresponding to the non-group descriptor in the non-group descriptor corresponding to the non-group descriptor may be a word frequency-inverse text frequency index of the non-group descriptor, and a component different from the component corresponding to the non-group descriptor in the word vector in the group descriptor is a first preset value (e.g., 0).
And step 304, training an initial deep learning model by taking the word vectors in the positive sample set and the negative sample set as actual input and taking the corresponding labeled classification result as expected output to obtain a group descriptor classification model.
Here, with the positive sample set and the negative sample set, the executing agent of the training step may train the initial deep learning model with word vectors in the positive sample set and the negative sample set as actual inputs and with corresponding labeled classification results as expected outputs, so as to obtain a group descriptor classification model. Specifically, the following can be performed:
first, the model structure of the initial deep learning model may be determined.
Here, the initial deep learning model may include various deep learning models. For example, the initial deep learning model may include at least one of: convolutional neural networks, cyclic neural networks, long-short term memory networks, conditional random fields.
By way of example, if the initial deep learning model is determined to be a convolutional neural network, it can be determined which layers the convolutional neural network specifically includes, such as which convolutional layers, pooling layers, fully-connected layers, and precedence relationships between layers. If convolutional layers are included, the size of the convolutional kernel of the convolutional layer, the convolution step size, can be determined. If a pooling layer is included, a pooling method may be determined.
Second, initial values of model parameters included in the initial deep learning model may be determined.
For example, if the initial deep learning model is determined to be a convolutional neural network, here, convolutional kernel parameters of convolutional layers that may be included in the convolutional neural network may be initialized, connection parameters for fully-connected layers may be initialized, and so on.
Finally, a parameter adjustment operation may be performed until a preset training end condition is satisfied, where the parameter adjustment operation may include: inputting a word vector corresponding to a group descriptor in a positive sample set or a word vector corresponding to a non-group descriptor in a negative sample set into an initial deep learning model, inputting the initial deep learning model to obtain a corresponding actual output result, calculating a difference between the actual output result and a corresponding labeling classification result for indicating that the group descriptor is the group descriptor or a labeling classification result for indicating the non-group descriptor, and adjusting a model parameter of the initial deep learning model based on the obtained difference. Here, the training end condition may include, for example, at least one of: and performing parameter adjustment operation for the positive sample set and the negative sample set, wherein the number of times of performing the parameter adjustment operation reaches a preset maximum training number, the calculated difference is smaller than a preset difference threshold value, and the parameter adjustment operation is performed on both the positive sample set and the negative sample set.
Through the parameter adjustment operation, the model parameters of the initial deep learning model are optimized, and the initial deep learning model after the parameter optimization can be determined as a group descriptor classification model. It should be noted that how to adjust and optimize the model parameters of the initial deep learning model based on the calculated differences is a prior art widely studied and applied in the field, and is not described herein again. For example, a gradient descent method may be employed.
By using the training steps shown in the above-mentioned flow 300, the group descriptor classification model can be automatically generated, and the labor cost for generating the group descriptor classification model is reduced. The expression of people changes along with the time, the reaction also changes in the alarm receiving text, and in addition, a novel group can also appear along with the development of the society. At this time, a new training sample set can be obtained, and the updated group descriptor classification model can be obtained by re-training through a training step so as to meet the requirement of the change of the expression mode of the current alarm receiving and processing text and the requirement of the extraction of the novel group descriptor.
The method provided by the embodiment of the disclosure inputs the word vector corresponding to each participle in the participle sequence obtained by segmenting the alarm receiving text to be recognized into the group descriptor classification model by using the group descriptor classification model trained in advance, so as to determine whether each participle is a group descriptor. And finally, generating a group description text set corresponding to the alarm receiving and processing text to be identified based on the determined classification result. Therefore, the group description text set corresponding to the alarm receiving and processing text to be identified is automatically generated, and the labor cost and the time cost are reduced. Furthermore, based on the obtained group description text, the reaction speed of the public security organization to the group and the group event can be improved, and the public security organization can accurately deal with the group and the group event.
With further reference to fig. 4, as an implementation of the methods shown in the above-mentioned figures, the present disclosure provides an embodiment of a deep learning model-based group identification apparatus, which corresponds to the embodiment of the method shown in fig. 2, and which can be applied in various electronic devices.
As shown in fig. 4, the deep learning model-based population recognition apparatus 400 of the present embodiment includes: word segmentation unit 401, classification unit 402, first generation unit 403, and second generation unit 404. The word segmentation unit 401 is configured to segment words of the alarm receiving and processing text to be identified to obtain a corresponding word segmentation sequence; a classification unit 402 configured to, for each participle in the obtained participle sequence, input a word vector corresponding to the participle into a group descriptor classification model to determine whether the participle is a group descriptor, wherein the group descriptor classification model is obtained by training in advance based on a deep learning model, and the group descriptor is a word in a group description text for describing a group; a first generating unit 403 configured to generate a group description text using a segmentation sequence segment composed of consecutive adjacent group description words in the obtained segmentation sequence; and the second generating unit 404 is configured to generate a group description text set corresponding to the alarm receiving and processing text to be identified by using each generated group description text.
In this embodiment, the specific processes of the word segmentation unit 401, the classification unit 402, the first generation unit 403, and the second generation unit 404 of the deep learning model-based population recognition apparatus 400 and the technical effects thereof can refer to the related descriptions of step 201, step 202, step 203, and step 204 in the corresponding embodiment of fig. 2, which are not repeated herein.
In some alternative embodiments, the population descriptor classification model may be obtained by pre-training through the following training steps: acquiring a training sample set, wherein the training sample comprises a word segmentation sequence obtained by segmenting a historical alarm receiving text and a corresponding labeling information sequence, and the labeling information in the labeling information sequence is used for indicating whether a corresponding word in the corresponding word segmentation sequence is a group descriptor in a group description text included in the corresponding historical alarm receiving text; determining a group descriptor and a non-group descriptor in the word segmentation sequence of each training sample of the training sample set according to the labeled information sequence in each training sample of the training sample set; generating a positive sample set and a negative sample set, wherein the positive sample comprises a word vector corresponding to the determined population descriptors and an annotation classification result used for indicating that the population descriptors are the population descriptors, and the negative sample comprises a word vector corresponding to the determined non-population descriptors and an annotation classification result used for indicating the non-population descriptors; and training an initial deep learning model by taking the word vectors in the positive sample set and the negative sample set as actual input and taking corresponding labeled classification results as expected output to obtain the group descriptor classification model.
In some optional embodiments, each component in the word vector corresponding to the group descriptor and each component in the word vector corresponding to the non-group descriptor may respectively correspond to each word in a preset dictionary in a one-to-one manner, a component in the word vector corresponding to the group descriptor in the group descriptor corresponding to the group descriptor may be a word frequency-inverse text frequency index TF-IDF of the group descriptor, a component different from the component corresponding to the group descriptor may be a first preset value, a component in the word vector corresponding to the non-group descriptor in the non-group descriptor corresponding to the non-group descriptor may be a word frequency-inverse text frequency index TF-IDF of the non-group descriptor, and a component different from the component corresponding to the non-group descriptor may be the first preset value.
In some alternative embodiments, a ratio of the number of positive samples in the positive sample set divided by the number of negative samples in the negative sample set may be within a preset ratio.
It should be noted that, for details of implementation and technical effects of each unit in the group identification device based on the deep learning model provided by the present disclosure, reference may be made to descriptions of other embodiments in the present disclosure, and details are not repeated herein.
Referring now to FIG. 5, a block diagram of a computer system 500 suitable for use in implementing the electronic device of the present disclosure is shown. The electronic device shown in fig. 5 is only an example, and should not bring any limitation to the functions and the scope of use of the present disclosure.
As shown in fig. 5, the computer system 500 includes a Central Processing Unit (CPU)501 that can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM) 502 or a program loaded from a storage section 508 into a Random Access Memory (RAM) 503. In the RAM 503, various programs and data necessary for the operation of the system 500 are also stored. The CPU 501, ROM 502, and RAM 503 are connected to each other via a bus 504. An Input/Output (I/O) interface 505 is also connected to bus 504.
The following components are connected to the I/O interface 505: an input section 506 including a touch screen, a tablet, a keyboard, a mouse, or the like; an output section 507 including a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, a speaker, and the like; a storage portion 508 including a hard disk and the like; and a communication section 509 including a Network interface card such as a LAN (Local Area Network) card, a modem, or the like. The communication section 509 performs communication processing via a network such as the internet.
In particular, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 509. The above-described functions defined in the method of the present disclosure are performed when the computer program is executed by a Central Processing Unit (CPU) 501. It should be noted that the computer readable medium in the present disclosure may be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having 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. In the present disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In contrast, in the present disclosure, a computer-readable signal medium may include a propagated data signal with computer-readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
Computer program code for carrying out operations for aspects of the present disclosure may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + +, Python, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The units described in this disclosure may be implemented by software or hardware. The described units may also be provided in a processor, and may be described as: a processor includes a word segmentation unit, a classification unit, a first generation unit, and a second generation unit. The names of the units do not form a limitation on the units themselves under certain conditions, for example, the word segmentation unit can also be described as a unit for segmenting the text of the alarm receiving and processing to be recognized into corresponding word segmentation sequences.
As another aspect, the present disclosure also provides a computer-readable medium, which may be contained in the apparatus described in the above embodiments; or may be present separately and not assembled into the device. The computer readable medium carries one or more programs which, when executed by the apparatus, cause the apparatus to: performing word segmentation on the alarm receiving and processing text to be recognized to obtain a corresponding word segmentation sequence; for each participle in the obtained participle sequence, inputting a word vector corresponding to the participle into a group descriptor classification model to determine whether the participle is a group descriptor or not, wherein the group descriptor classification model is obtained by pre-training based on a deep learning model, and the group descriptor is a word in a group description text for describing a group; generating a group description text by using a segmentation sequence segment consisting of continuous adjacent group description words in the obtained segmentation sequence; and generating a group description text set corresponding to the alarm receiving and processing text to be identified by using the generated group description texts.
The foregoing description is only exemplary of the preferred embodiments of the disclosure and is illustrative of the principles of the technology employed. It will be appreciated by those skilled in the art that the scope of the invention in the present disclosure is not limited to the specific combination of the above-mentioned features, but also encompasses other embodiments in which any combination of the above-mentioned features or their equivalents is possible without departing from the inventive concept as defined above. For example, the above features and (but not limited to) the features disclosed in this disclosure having similar functions are replaced with each other to form the technical solution.
Claims (10)
1. A population identification method based on a deep learning model comprises the following steps:
performing word segmentation on the alarm receiving and processing text to be recognized to obtain a corresponding word segmentation sequence;
for each participle in the obtained participle sequence, inputting a word vector corresponding to the participle into a group descriptor classification model to determine whether the participle is a group descriptor or not, wherein the group descriptor classification model is obtained by pre-training based on a deep learning model, and the group descriptor is a word in a group description text for describing a group;
generating a group description text by using a segmentation sequence segment consisting of continuous adjacent group description words in the obtained segmentation sequence;
and generating a group description text set corresponding to the alarm receiving and processing text to be identified by using each generated group description text.
2. The method of claim 1, wherein the population descriptor classification model is pre-trained by the following training steps:
acquiring a training sample set, wherein the training sample comprises a segmentation sequence obtained by segmenting a historical alarm receiving and processing text and a corresponding labeling information sequence, and the labeling information in the labeling information sequence is used for indicating whether a corresponding segmentation in the corresponding segmentation sequence belongs to a group description text included in the corresponding historical alarm receiving and processing text;
determining a group descriptor and a non-group descriptor in the word segmentation sequence of each training sample of the training sample set according to the labeling information sequence in each training sample of the training sample set;
generating a positive sample set and a negative sample set, wherein the positive sample comprises a word vector corresponding to the determined population descriptors and an annotation classification result used for indicating that the population descriptors are the population descriptors, and the negative sample comprises a word vector corresponding to the determined non-population descriptors and an annotation classification result used for indicating the non-population descriptors;
and training an initial deep learning model by taking the word vectors in the positive sample set and the negative sample set as actual input and taking corresponding labeling classification results as expected output to obtain the population descriptor classification model.
3. The method according to claim 2, wherein each component in the word vector corresponding to the group descriptor and each component in the word vector corresponding to the non-group descriptor respectively correspond to each word in a preset dictionary in a one-to-one manner, the component corresponding to the group descriptor in the word vector corresponding to the group descriptor in the group descriptor is a word frequency-inverse text frequency index TF-IDF of the group descriptor, the component different from the component corresponding to the group descriptor is a first preset value, the component corresponding to the non-group descriptor in the word vector corresponding to the non-group descriptor in the non-group descriptor is the word frequency-inverse text frequency index TF-IDF of the non-group descriptor, and the component different from the component corresponding to the non-group descriptor is the first preset value.
4. The method of claim 3, wherein a ratio of the number of positive samples in the set of positive samples divided by the number of negative samples in the set of negative samples is within a preset ratio.
5. A deep learning model-based population recognition apparatus, comprising:
the word segmentation unit is configured to segment words of the alarm receiving and processing text to be identified to obtain a corresponding word segmentation sequence;
a classification unit configured to, for each participle in the obtained participle sequence, input a word vector corresponding to the participle into a group descriptor classification model to determine whether the participle is a group descriptor, wherein the group descriptor classification model is obtained by pre-training based on a deep learning model, and the group descriptor is a word in a group description text for describing a group;
a first generating unit configured to generate a group description text using a segmentation sequence segment composed of consecutive adjacent group description words in the obtained segmentation sequence;
and the second generating unit is configured to generate a group description text set corresponding to the alarm receiving and processing text to be identified by using each generated group description text.
6. The apparatus of claim 5, wherein the population descriptor classification model is pre-trained by the following training steps:
acquiring a training sample set, wherein the training sample comprises a word segmentation sequence obtained by segmenting a historical alarm receiving text and a corresponding labeling information sequence, and the labeling information in the labeling information sequence is used for indicating whether a corresponding word in the corresponding word segmentation sequence is a group descriptor in a group description text included in the corresponding historical alarm receiving text;
determining a group descriptor and a non-group descriptor in the word segmentation sequence of each training sample of the training sample set according to the labeling information sequence in each training sample of the training sample set;
generating a positive sample set and a negative sample set, wherein the positive sample comprises a word vector corresponding to the determined population descriptors and an annotation classification result used for indicating that the population descriptors are the population descriptors, and the negative sample comprises a word vector corresponding to the determined non-population descriptors and an annotation classification result used for indicating the non-population descriptors;
and training an initial deep learning model by taking the word vectors in the positive sample set and the negative sample set as actual input and taking corresponding labeling classification results as expected output to obtain the population descriptor classification model.
7. The apparatus according to claim 6, wherein components in the word vectors corresponding to the group descriptors and the word vectors corresponding to the non-group descriptors are respectively in one-to-one correspondence with words in a preset dictionary, the component corresponding to the group descriptors in the word vectors corresponding to the group descriptors is a word frequency-inverse text frequency index TF-IDF of the group descriptors, the component different from the component corresponding to the group descriptors is a first preset value, the component corresponding to the non-group descriptors in the word vectors corresponding to the non-group descriptors is a word frequency-inverse text frequency index TF-IDF of the non-group descriptors, and the component different from the component corresponding to the non-group descriptors is the first preset value.
8. The apparatus of claim 7, wherein a ratio of the number of positive samples in the set of positive samples divided by the number of negative samples in the set of negative samples is within a preset ratio.
9. An electronic device, comprising:
one or more processors;
storage means for storing one or more programs;
the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method recited in any of claims 1-4.
10. A computer-readable medium, on which a computer program is stored, wherein the program, when executed by a processor, implements the method of any one of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010348490.XA CN113111174A (en) | 2020-04-28 | 2020-04-28 | Group identification method, device, equipment and medium based on deep learning model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010348490.XA CN113111174A (en) | 2020-04-28 | 2020-04-28 | Group identification method, device, equipment and medium based on deep learning model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113111174A true CN113111174A (en) | 2021-07-13 |
Family
ID=76709181
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010348490.XA Pending CN113111174A (en) | 2020-04-28 | 2020-04-28 | Group identification method, device, equipment and medium based on deep learning model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113111174A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113743110A (en) * | 2021-11-08 | 2021-12-03 | 京华信息科技股份有限公司 | Word missing detection method and system based on fine-tuning generation type confrontation network model |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106095996A (en) * | 2016-06-22 | 2016-11-09 | 量子云未来（北京）信息科技有限公司 | Method for text classification |
| CN106294684A (en) * | 2016-08-06 | 2017-01-04 | 上海高欣计算机系统有限公司 | The file classification method of term vector and terminal unit |
| CN108009228A (en) * | 2017-11-27 | 2018-05-08 | 咪咕互动娱乐有限公司 | A kind of method to set up of content tab, device and storage medium |
| CN109271574A (en) * | 2018-08-28 | 2019-01-25 | 麒麟合盛网络技术股份有限公司 | A kind of hot word recommended method and device |
| CN109684478A (en) * | 2018-12-18 | 2019-04-26 | 腾讯科技（深圳）有限公司 | Disaggregated model training method, classification method and device, equipment and medium |
| CN110795553A (en) * | 2019-09-09 | 2020-02-14 | 腾讯科技（深圳）有限公司 | Abstract generation method and device |
| CN110909164A (en) * | 2019-11-22 | 2020-03-24 | 科大国创软件股份有限公司 | Text enhancement semantic classification method and system based on convolutional neural network |
| CN110991171A (en) * | 2019-09-30 | 2020-04-10 | 奇安信科技集团股份有限公司 | Sensitive word detection method and device |
-
2020
- 2020-04-28 CN CN202010348490.XA patent/CN113111174A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106095996A (en) * | 2016-06-22 | 2016-11-09 | 量子云未来（北京）信息科技有限公司 | Method for text classification |
| CN106294684A (en) * | 2016-08-06 | 2017-01-04 | 上海高欣计算机系统有限公司 | The file classification method of term vector and terminal unit |
| CN108009228A (en) * | 2017-11-27 | 2018-05-08 | 咪咕互动娱乐有限公司 | A kind of method to set up of content tab, device and storage medium |
| CN109271574A (en) * | 2018-08-28 | 2019-01-25 | 麒麟合盛网络技术股份有限公司 | A kind of hot word recommended method and device |
| CN109684478A (en) * | 2018-12-18 | 2019-04-26 | 腾讯科技（深圳）有限公司 | Disaggregated model training method, classification method and device, equipment and medium |
| CN110795553A (en) * | 2019-09-09 | 2020-02-14 | 腾讯科技（深圳）有限公司 | Abstract generation method and device |
| CN110991171A (en) * | 2019-09-30 | 2020-04-10 | 奇安信科技集团股份有限公司 | Sensitive word detection method and device |
| CN110909164A (en) * | 2019-11-22 | 2020-03-24 | 科大国创软件股份有限公司 | Text enhancement semantic classification method and system based on convolutional neural network |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113743110A (en) * | 2021-11-08 | 2021-12-03 | 京华信息科技股份有限公司 | Word missing detection method and system based on fine-tuning generation type confrontation network model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109104620B (en) | Short video recommendation method and device and readable medium | |
| CN108427939B (en) | Model generation method and device | |
| US20210019531A1 (en) | Method and apparatus for classifying video | |
| CN108121699B (en) | Method and apparatus for outputting information | |
| CN110633423B (en) | Target account identification method, device, equipment and storage medium | |
| CN113657113A (en) | Text processing method and device and electronic equipment | |
| CN111723180A (en) | Interviewing method and device | |
| US20210349920A1 (en) | Method and apparatus for outputting information | |
| CN113033707B (en) | Video classification method and device, readable medium and electronic equipment | |
| CN113051911B (en) | Method, apparatus, device, medium and program product for extracting sensitive words | |
| CN113111167A (en) | Method and device for extracting vehicle model of alarm receiving and processing text based on deep learning model | |
| CN113111174A (en) | Group identification method, device, equipment and medium based on deep learning model | |
| CN112633004A (en) | Text punctuation deletion method and device, electronic equipment and storage medium | |
| CN113255819B (en) | Method and device for identifying information | |
| CN113111169A (en) | Deep learning model-based alarm receiving and processing text address information extraction method and device | |
| CN113111165A (en) | Deep learning model-based alarm receiving warning condition category determination method and device | |
| CN113742450A (en) | User data grade label falling method and device, electronic equipment and storage medium | |
| CN114020896A (en) | Intelligent question and answer method, system, electronic equipment and storage medium | |
| CN113111897A (en) | Alarm receiving and warning condition type determining method and device based on support vector machine | |
| CN113920397A (en) | Method and device for training image classification model and method and device for image classification | |
| CN113094499A (en) | Deep learning model-based organization identification method and device, equipment and medium | |
| CN113111234A (en) | Regular expression-based alarm condition category determination method and device | |
| CN113111175A (en) | Extreme behavior identification method, device, equipment and medium based on deep learning model | |
| CN113111171A (en) | Deep learning model-based alarm handling and warning condition category determination method and device | |
| CN113111176A (en) | Deep learning model based crime solution identification method and device, equipment and medium |
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 |