CN111061870B - Article quality evaluation method and device - Google Patents
Article quality evaluation method and device Download PDFInfo
- Publication number
- CN111061870B CN111061870B CN201911163934.6A CN201911163934A CN111061870B CN 111061870 B CN111061870 B CN 111061870B CN 201911163934 A CN201911163934 A CN 201911163934A CN 111061870 B CN111061870 B CN 111061870B
- Authority
- CN
- China
- Prior art keywords
- article
- semantic feature
- neural network
- evaluated
- network model
- 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.)
- Active
Links
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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Abstract
The embodiment of the disclosure discloses a quality evaluation method and a quality evaluation device for articles, relates to the technical field of computers, and can solve the technical problems that most news manuscripts in the prior art have timeliness and are difficult to complete auditing in time by a small amount of manpower. The method of the embodiment of the disclosure mainly comprises the following steps: calculating semantic feature vectors of articles to be evaluated according to the first neural network model; classifying the semantic feature vectors according to a second neural network model, and evaluating to obtain the quality of the article to be evaluated. Compared with the prior art, the method has the advantages that manual auditing is not needed in the process of evaluating the quality of the articles to be evaluated, automation of article quality evaluation is realized, the evaluation efficiency of article quality can be greatly improved, and mass quality evaluation work of articles to be evaluated can be completed in time.
Description
Technical Field
The embodiment of the disclosure relates to the technical field of computers, in particular to a method and a device for evaluating the quality of an article.
Background
With the popularization of the internet of things, more common people are added into the lines and rows of the internet publication.
Taking the news industry as an example, because of the limited number of professional reporters, with the increasing demand for web news, more free contributors are added to the line of news contributors. Most of the free contributors do not have a training experience of professional news manuscript writing, so that the problem of uneven quality exists in the manuscripts written by the free contributors. The problems greatly increase the workload of news contributors, lead to a plurality of news contributors and future and serious audits, and then send the news contributors to the network for news release, so that a large number of news contributors with lower quality exist on the network, and poor reading experience is brought to users.
Disclosure of Invention
The method and the device solve the technical problems that the number of the to-be-checked manuscripts is huge, the timeliness exists in most news manuscripts, and the checking is difficult to complete in time by a small amount of manpower.
The embodiment of the disclosure mainly provides the following technical scheme:
in a first aspect, an embodiment of the present disclosure provides a method for evaluating quality of an article, including:
calculating semantic feature vectors of articles to be evaluated according to the first neural network model;
classifying the semantic feature vectors according to a second neural network model, and evaluating to obtain the quality of the article to be evaluated.
In some embodiments, before classifying the semantic feature vectors according to a second neural network model, the method further comprises:
calculating semantic feature vectors of articles to be trained according to the first neural network model;
and performing model training according to the semantic feature vector of the article to be trained and the quality label of the article to be trained to obtain the second neural network model.
In some embodiments, computing semantic feature vectors of articles to be evaluated from a first neural network model includes:
acquiring semantic feature vectors corresponding to the characters of the article to be evaluated according to the first neural network model;
and calculating the semantic feature vector of the article to be evaluated according to the semantic feature vector corresponding to each word in the article to be evaluated.
In some embodiments, computing semantic feature vectors of articles to be evaluated from a first neural network model includes:
and obtaining the weight of each word through the universal semantic feature vector and the semantic feature vector of each word of the current sentence, and calculating the semantic feature vector of the article to be evaluated through the weight and the semantic feature vector of each word.
In some embodiments, classifying the semantic feature vectors according to a second neural network model includes:
classifying the articles to be evaluated according to a third neural network model;
and selecting a second neural network model corresponding to the category to which the article to be evaluated belongs, classifying the semantic feature vector, and evaluating to obtain the quality of the article to be evaluated.
In a second aspect, embodiments of the present disclosure provide a quality evaluation apparatus for an article, including:
the computing unit is used for computing semantic feature vectors of articles to be evaluated according to the first neural network model;
and the evaluation unit is used for classifying the semantic feature vectors according to the second neural network model and evaluating to obtain the quality of the article to be evaluated.
In some embodiments, the computing unit is further configured to compute a semantic feature vector of an article to be trained according to the first neural network model;
and the model training unit is used for carrying out model training according to the semantic feature vector of the article to be trained and the quality label of the article to be trained to obtain the second neural network model.
In some embodiments, the computing unit comprises:
the acquisition module is used for acquiring semantic feature vectors corresponding to the characters of the article to be evaluated according to the first neural network model;
and the first calculation module is used for calculating the semantic feature vector of the article to be evaluated according to the semantic feature vector corresponding to each word in the article to be evaluated.
In some embodiments, the computing unit comprises:
and the second calculation module is used for obtaining the weight of each word through the universal semantic feature vector and the semantic feature vector of each word of the current sentence, and calculating the semantic feature vector of the article to be evaluated through the weight and the semantic feature vector of each word.
In some embodiments, the evaluation unit comprises:
the classification module is used for classifying the articles to be evaluated according to a third neural network model;
and the evaluation module is used for selecting a second neural network model corresponding to the category to which the article to be evaluated belongs, classifying the semantic feature vectors, and evaluating to obtain the quality of the article to be evaluated.
In a third aspect, an embodiment of the present disclosure provides a storage medium, where the storage medium includes a stored program, and when the program runs, controls a device in which the storage medium is located to execute the method for evaluating quality of the article in the first aspect.
The storage medium may include volatile memory, random Access Memory (RAM) and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM), among other forms in computer readable media, the memory including at least one memory chip.
In a fourth aspect, embodiments of the present disclosure provide an apparatus for evaluating quality of an article of text, the apparatus comprising a storage medium; and one or more processors coupled to the storage medium, the processors configured to execute the program instructions stored in the storage medium; the program instructions, when run, perform the method of evaluating quality of an article according to the first aspect.
By means of the technical scheme, the article quality evaluation method and device provided by the technical scheme of the invention have at least the following advantages:
in the technical scheme provided by the embodiment of the disclosure, after writing the article to be evaluated, the manuscript staff can calculate the semantic feature vector through the first neural network model, then input the calculated semantic feature vector into the second neural network model to obtain a classification result, and evaluate the quality of the article to be evaluated according to the classification result. Compared with the prior art, the method has the advantages that manual auditing is not needed in the process of evaluating the quality of the articles to be evaluated, automation of article quality evaluation is realized, the evaluation efficiency of article quality can be greatly improved, and mass quality evaluation work of articles to be evaluated can be completed in time.
The foregoing description is merely an overview of the technical solutions of the embodiments of the present disclosure, and may be implemented according to the content of the specification in order to make the technical means of the embodiments of the present disclosure more clearly understood, and in order to make the foregoing and other objects, features and advantages of the embodiments of the present disclosure more comprehensible, the following detailed description of the embodiments of the present disclosure.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the embodiments of the disclosure. Also, like reference numerals are used to designate like parts throughout the figures. In the drawings:
FIG. 1 illustrates a flow chart of a method of quality assessment of an article provided by an embodiment of the present disclosure;
FIG. 2 illustrates a flow chart of another article quality evaluation method provided by an embodiment of the present disclosure;
FIG. 3 shows a block diagram of a device for evaluating quality of an article provided by an embodiment of the present disclosure;
fig. 4 shows a block diagram of a specific article quality evaluation apparatus provided by an embodiment of the present disclosure.
Detailed Description
Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
In a first aspect, an embodiment of the present disclosure provides a method for evaluating quality of an article, as shown in fig. 1, where the method includes:
110. calculating semantic feature vectors of articles to be evaluated according to the first neural network model;
the first neural network model is a model for calculating semantic feature vectors of articles, and the first neural network model can calculate the semantic feature vectors of the articles to be evaluated by inputting the articles to be evaluated into the first neural network model.
In some implementations, the first neural network model is a model for acquiring semantic feature vectors of words in the article to be evaluated, and calculating the semantic feature vectors of the article to be evaluated according to the first neural network model includes: acquiring semantic feature vectors corresponding to the characters of the article to be evaluated according to the first neural network model; and calculating the semantic feature vector of the article to be evaluated according to the semantic feature vector corresponding to each word in the article to be evaluated. The calculating the semantic feature vector of the article to be evaluated according to the semantic feature vector corresponding to each word in the article to be evaluated may adopt an average calculating mode, including: and accumulating and averaging the semantic feature vectors of each text to obtain the semantic feature vector of the article to be evaluated.
In some implementations, the semantic feature vector of the article to be evaluated is calculated according to the first neural network model, and a weighted model calculation manner may be adopted, including: and obtaining the weight of each word through the universal semantic feature vector and the semantic feature vector of each word of the current sentence, and calculating the semantic feature vector of the article to be evaluated through the weight and the semantic feature vector of each word. Specifically, in the weighted calculation, a transducer model structure may be used to obtain a semantic feature vector of the first text according to the semantic feature vector of each word. The transition model based on the attribute abandons the inherent definite form and does not use any CNN or RNN structure, and the model can work in high parallel and has higher semantic analysis performance.
In general, articles to be evaluated may be categorized into different categories, such as news documents, web novels, real-time comments, etc., with different categories corresponding to different first neural network models. Of course, the articles to be evaluated may not be classified, and embodiments of the present application are not limited.
The semantic feature vector of the article to be evaluated can be calculated in real time by adopting the first neural network model, and compared with the artificial quality evaluation method, the quality evaluation of the article with higher timeliness can be realized.
120. Classifying the semantic feature vectors according to a second neural network model, and evaluating to obtain the quality of the article to be evaluated.
The second neural network model is a classification model for calculating the quality corresponding to the semantic feature vector of the article to be evaluated, and by inputting the semantic feature vector into the second neural network model, the second neural network model can calculate the quality corresponding to the semantic feature vector, namely, evaluate the quality of the article to be evaluated. Generally, the quality of the article to be evaluated may include pass and fail, and it is easy to understand that the quality of the article to be evaluated may be set in any other form according to different requirements, for example, the quality of the article to be evaluated may include good, bad, and poor.
In the technical scheme provided by the embodiment of the disclosure, after writing the article to be evaluated, the manuscript staff can calculate the semantic feature vector through the first neural network model, then input the calculated semantic feature vector into the second neural network model to obtain a classification result, and evaluate the quality of the article to be evaluated according to the classification result. Compared with the prior art, the method has the advantages that manual auditing is not needed in the process of evaluating the quality of the articles to be evaluated, automation of article quality evaluation is realized, the evaluation efficiency of article quality can be greatly improved, and mass quality evaluation work of articles to be evaluated can be completed in time.
In a second aspect, an embodiment of the present disclosure provides a method for evaluating quality of an article, in which a step of training a second neural network model in the method for evaluating quality of an article of the first aspect is disclosed, as shown in fig. 2, the method includes:
210. calculating semantic feature vectors of articles to be trained according to the first neural network model;
the first neural network model is a model for calculating semantic feature vectors of articles, and the first neural network model can calculate the semantic feature vectors of the articles to be trained by inputting the articles to be trained into the first neural network model.
In some implementations, the first neural network model is a model for acquiring semantic feature vectors of Chinese characters in an article, and calculating the semantic feature vectors of the article to be trained according to the first neural network model includes: acquiring semantic feature vectors corresponding to the characters of the article to be trained according to the first neural network model; and calculating the semantic feature vector of the article to be trained according to the semantic feature vector corresponding to each word in the article to be trained. The calculating the semantic feature vector of the article to be trained according to the semantic feature vector corresponding to each word in the article to be trained may adopt an average calculating mode, including: and accumulating and averaging the semantic feature vectors of each text to obtain the semantic feature vector of the article to be trained.
In some implementations, the semantic feature vectors of the articles to be trained are calculated according to the first neural network model, and a weighted model calculation mode can be adopted, including: and obtaining the weight of each word through the universal semantic feature vector and the semantic feature vector of each word of the current sentence, and calculating the semantic feature vector of the article to be trained through the weight and the semantic feature vector of each word. Specifically, in the weighted calculation, a transducer model structure can be adopted to obtain the semantic feature vector of the first text according to the semantic feature vector of each word, the transducer model based on the Attention abandons the inherent fixed form and does not use any CNN or RNN structure, and the model can work in high parallel and has higher semantic analysis performance.
In general, articles to be trained may be classified into different categories, such as news documents, web novels, real-time comments, etc., with the different categories corresponding to different first neural network models. Of course, the articles to be trained may not be classified, and embodiments of the present application are not limited. The semantic feature vector of the article to be trained can be calculated in real time by adopting the first neural network model.
220. And performing model training according to the semantic feature vector of the article to be trained and the quality label of the article to be trained to obtain the second neural network model.
A large number of articles to be trained can be collected for each quality of articles, semantic feature vectors (X) corresponding to each article are calculated in step 210, each article to be trained is labeled, and the labels mark the quality (Y) corresponding to the article to be trained, so that a training data set (X, Y) is formed. The neural network classification model is trained by the training dataset (X, Y). The trained neural network classification model is a second neural network model which can evaluate the quality of the articles to be evaluated.
In implementation, for articles to be trained with different classifications, model training can be performed according to semantic feature vectors of the articles to be trained with different classifications and quality labels of the articles to be trained with different classifications, so as to obtain second neural network models with different classifications.
230. Calculating semantic feature vectors of articles to be evaluated according to the first neural network model;
240. classifying the semantic feature vectors according to a second neural network model, and evaluating to obtain the quality of the article to be evaluated.
In implementation, for articles to be trained with different classifications, second neural network models corresponding to the classifications with different classifications are provided, and the articles to be evaluated can be classified according to a third neural network model; and selecting a second neural network model corresponding to the category to which the article to be evaluated belongs, classifying the semantic feature vector, and evaluating to obtain the quality of the article to be evaluated. Under the condition that the categories of the articles to be evaluated are wider, the quality evaluation of the articles to be evaluated can be performed by adopting second neural network models of different categories.
In a third aspect, an embodiment of the present disclosure provides an article quality evaluation apparatus, as shown in fig. 3, the apparatus includes:
a calculating unit 10, configured to calculate a semantic feature vector of an article to be evaluated according to the first neural network model;
and the evaluation unit 20 is used for classifying the semantic feature vectors according to the second neural network model and evaluating to obtain the quality of the article to be evaluated.
In some embodiments, as shown in fig. 4, the calculating unit 10 is further configured to calculate a semantic feature vector of the article to be trained according to the first neural network model;
the model training unit 30 is configured to perform model training according to the semantic feature vector of the article to be trained and the quality label of the article to be trained, so as to obtain the second neural network model.
In some embodiments, the computing unit 10 includes:
the obtaining module 11 is configured to obtain, according to the first neural network model, a semantic feature vector corresponding to a text of the article to be evaluated;
the first calculating module 12 is configured to calculate a semantic feature vector of the article to be evaluated according to the semantic feature vector corresponding to each text in the article to be evaluated.
In some embodiments, the computing unit 10 includes:
and the second calculating module 13 is used for obtaining the weight of each word through the universal semantic feature vector and the semantic feature vector of each word of the current sentence, and calculating the semantic feature vector of the article to be evaluated through the weight and the semantic feature vector of each word.
In some embodiments, the evaluation unit 20 comprises:
a classification module 21, configured to classify the articles to be evaluated according to a third neural network model;
and the evaluation module 22 is used for selecting a second neural network model corresponding to the category to which the article to be evaluated belongs, classifying the semantic feature vectors, and evaluating to obtain the quality of the article to be evaluated.
In a fourth aspect, an embodiment of the present disclosure provides a storage medium, where the storage medium includes a stored program, and when the program runs, controls a device in which the storage medium is located to execute the method for evaluating quality of an article in the first aspect or the second aspect.
The storage medium may include volatile memory, random Access Memory (RAM) and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM), among other forms in computer readable media, the memory including at least one memory chip.
In a fifth aspect, embodiments of the present disclosure provide an apparatus for evaluating quality of an article, the apparatus comprising a storage medium; and one or more processors coupled to the storage medium, the processors configured to execute the program instructions stored in the storage medium; the program instructions, when executed, perform the method for evaluating the quality of an article according to the first or second aspect.
In a sixth aspect, A1, a method for evaluating quality of an article, includes:
calculating semantic feature vectors of articles to be evaluated according to the first neural network model;
classifying the semantic feature vectors according to a second neural network model, and evaluating to obtain the quality of the article to be evaluated.
A2, the article quality evaluation method according to A1, before classifying the semantic feature vectors according to a second neural network model, the method further comprises:
calculating semantic feature vectors of articles to be trained according to the first neural network model;
and performing model training according to the semantic feature vector of the article to be trained and the quality label of the article to be trained to obtain the second neural network model.
A3, calculating semantic feature vectors of the articles to be evaluated according to the quality evaluation method of the articles according to the A1, wherein the semantic feature vectors comprise:
acquiring semantic feature vectors corresponding to the characters of the article to be evaluated according to the first neural network model;
and calculating the semantic feature vector of the article to be evaluated according to the semantic feature vector corresponding to each word in the article to be evaluated.
A4, calculating semantic feature vectors of the articles to be evaluated according to the quality evaluation method of the articles according to the A1, wherein the semantic feature vectors comprise:
and obtaining the weight of each word through the universal semantic feature vector and the semantic feature vector of each word of the current sentence, and calculating the semantic feature vector of the article to be evaluated through the weight and the semantic feature vector of each word.
A5, classifying the semantic feature vectors according to a second neural network model according to the method for evaluating the quality of the article according to any one of A1-4, wherein the method comprises the following steps:
classifying the articles to be evaluated according to a third neural network model;
and selecting a second neural network model corresponding to the category to which the article to be evaluated belongs, classifying the semantic feature vector, and evaluating to obtain the quality of the article to be evaluated.
In a seventh aspect, B6 is an article quality evaluation device, comprising:
the computing unit is used for computing semantic feature vectors of articles to be evaluated according to the first neural network model;
and the evaluation unit is used for classifying the semantic feature vectors according to the second neural network model and evaluating to obtain the quality of the article to be evaluated.
B7, the quality evaluation device of the article according to B6,
the computing unit is also used for computing semantic feature vectors of articles to be trained according to the first neural network model;
and the model training unit is used for carrying out model training according to the semantic feature vector of the article to be trained and the quality label of the article to be trained to obtain the second neural network model.
B8, the article quality evaluation device according to B6, wherein the calculation unit includes:
the acquisition module is used for acquiring semantic feature vectors corresponding to the characters of the article to be evaluated according to the first neural network model;
and the first calculation module is used for calculating the semantic feature vector of the article to be evaluated according to the semantic feature vector corresponding to each word in the article to be evaluated.
B9, the article quality evaluation device according to B6, wherein the calculation unit includes:
and the second calculation module is used for obtaining the weight of each word through the universal semantic feature vector and the semantic feature vector of each word of the current sentence, and calculating the semantic feature vector of the article to be evaluated through the weight and the semantic feature vector of each word.
B10, the quality evaluation device of the article according to any one of B6 to B9, the evaluation unit including:
the classification module is used for classifying the articles to be evaluated according to a third neural network model;
and the evaluation module is used for selecting a second neural network model corresponding to the category to which the article to be evaluated belongs, classifying the semantic feature vectors, and evaluating to obtain the quality of the article to be evaluated.
In an eighth aspect, C11 is a storage medium, the storage medium including a stored program, wherein the program, when executed, controls a device in which the storage medium is located to execute the method for evaluating the quality of the article of any one of A1 to A5.
In a ninth aspect, D12 is an article quality evaluation device, the device comprising a storage medium; and one or more processors coupled to the storage medium, the processors configured to execute the program instructions stored in the storage medium; the program instructions, when run, perform a method of evaluating the quality of an article of any one of A1 to A5.
It will be apparent to those skilled in the art that embodiments of the present disclosure may be provided as a method, system, or computer program product. Accordingly, embodiments of the present disclosure may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Moreover, embodiments of the present disclosure may take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the disclosure. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, etc., such as Read Only Memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.
Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises an element.
It will be appreciated by those skilled in the art that embodiments of the present disclosure may be provided as a method, system, or computer program product. Accordingly, embodiments of the present disclosure may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Moreover, embodiments of the present disclosure may take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and changes may be made to the present application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. which are within the spirit and principles of the present application are intended to be included within the scope of the claims of the present application.
Claims (10)
1. A method for evaluating quality of an article, comprising:
calculating semantic feature vectors of articles to be evaluated according to the first neural network model;
classifying the articles to be evaluated according to a third neural network model;
selecting a second neural network model corresponding to the category to which the article to be evaluated belongs, classifying the semantic feature vector, and evaluating to obtain the quality of the article to be evaluated; the second neural network model is obtained by model training based on semantic feature vectors of articles to be trained and quality labels of the articles to be trained.
2. The method for evaluating the quality of an article according to claim 1, wherein the process of performing model training based on the semantic feature vector of the article to be trained and the quality label of the article to be trained to obtain the second neural network model comprises:
calculating semantic feature vectors of articles to be trained according to the first neural network model;
and performing model training according to the semantic feature vector of the article to be trained and the quality label of the article to be trained to obtain the second neural network model.
3. The method of claim 1, wherein calculating semantic feature vectors of articles to be evaluated based on a first neural network model comprises:
acquiring semantic feature vectors corresponding to the characters of the article to be evaluated according to the first neural network model;
and calculating the semantic feature vector of the article to be evaluated according to the semantic feature vector corresponding to each word in the article to be evaluated.
4. The method of claim 1, wherein calculating semantic feature vectors of articles to be evaluated based on a first neural network model comprises:
and obtaining the weight of each word through the universal semantic feature vector and the semantic feature vector of each word of the current sentence, and calculating the semantic feature vector of the article to be evaluated through the weight and the semantic feature vector of each word.
5. An article quality evaluation device comprising:
the computing unit is used for computing semantic feature vectors of articles to be evaluated according to the first neural network model;
the evaluation unit is used for classifying the articles to be evaluated according to a third neural network model; selecting a second neural network model corresponding to the category to which the article to be evaluated belongs, classifying the semantic feature vector, and evaluating to obtain the quality of the article to be evaluated; the second neural network model is obtained by model training based on semantic feature vectors of articles to be trained and quality labels of the articles to be trained.
6. The apparatus for evaluating the quality of an article according to claim 5, wherein,
the computing unit is also used for computing semantic feature vectors of articles to be trained according to the first neural network model;
and the model training unit is used for carrying out model training according to the semantic feature vector of the article to be trained and the quality label of the article to be trained to obtain the second neural network model.
7. The article quality evaluation device according to claim 5, wherein the calculation unit includes:
the acquisition module is used for acquiring semantic feature vectors corresponding to the characters of the article to be evaluated according to the first neural network model;
and the first calculation module is used for calculating the semantic feature vector of the article to be evaluated according to the semantic feature vector corresponding to each word in the article to be evaluated.
8. The article quality evaluation device according to claim 5, wherein the calculation unit includes:
and the second calculation module is used for obtaining the weight of each word through the universal semantic feature vector and the semantic feature vector of each word of the current sentence, and calculating the semantic feature vector of the article to be evaluated through the weight and the semantic feature vector of each word.
9. A storage medium comprising a stored program, wherein the program, when run, controls a device in which the storage medium is located to perform the quality evaluation method of the article of any one of claims 1 to 4.
10. An article quality evaluation device, characterized in that the device comprises a storage medium; and one or more processors coupled to the storage medium, the processors configured to execute the program instructions stored in the storage medium; the program instructions, when executed, perform the method of quality assessment of an article of any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911163934.6A CN111061870B (en) | 2019-11-25 | 2019-11-25 | Article quality evaluation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911163934.6A CN111061870B (en) | 2019-11-25 | 2019-11-25 | Article quality evaluation method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111061870A CN111061870A (en) | 2020-04-24 |
| CN111061870B true CN111061870B (en) | 2023-06-06 |
Family
ID=70298181
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911163934.6A Active CN111061870B (en) | 2019-11-25 | 2019-11-25 | Article quality evaluation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111061870B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111523322A (en) * | 2020-04-25 | 2020-08-11 | 中信银行股份有限公司 | Requirement document quality evaluation model training method and requirement document quality evaluation method |
| CN111737446B (en) * | 2020-06-22 | 2024-04-05 | 北京百度网讯科技有限公司 | Method, apparatus, device and storage medium for constructing quality assessment model |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8234285B1 (en) * | 2009-07-10 | 2012-07-31 | Google Inc. | Context-dependent similarity measurements |
| CN110263350A (en) * | 2019-03-08 | 2019-09-20 | 腾讯科技（深圳）有限公司 | Model training method, device, computer readable storage medium and computer equipment |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104049755B (en) * | 2014-06-18 | 2017-01-18 | 中国科学院自动化研究所 | Information processing method and device |
| CN105589941A (en) * | 2015-12-15 | 2016-05-18 | 北京百分点信息科技有限公司 | Emotional information detection method and apparatus for web text |
| CN107506360B (en) * | 2016-06-14 | 2020-09-11 | 科大讯飞股份有限公司 | Article scoring method and system |
| CN107133211B (en) * | 2017-04-26 | 2020-06-16 | 中国人民大学 | Composition scoring method based on attention mechanism |
| CN108073571B (en) * | 2018-01-12 | 2021-08-13 | 中译语通科技股份有限公司 | Multi-language text quality evaluation method and system and intelligent text processing system |
| CN108519975B (en) * | 2018-04-03 | 2021-09-28 | 北京先声教育科技有限公司 | Composition scoring method, device and storage medium |
| CN109492157B (en) * | 2018-10-24 | 2021-08-31 | 华侨大学 | News recommendation method and theme characterization method based on RNN and attention mechanism |
| CN110188350B (en) * | 2019-05-22 | 2021-06-01 | 北京百度网讯科技有限公司 | Text consistency calculation method and device |
| CN110162797B (en) * | 2019-06-21 | 2023-04-07 | 北京百度网讯科技有限公司 | Article quality detection method and device |
-
2019
- 2019-11-25 CN CN201911163934.6A patent/CN111061870B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8234285B1 (en) * | 2009-07-10 | 2012-07-31 | Google Inc. | Context-dependent similarity measurements |
| CN110263350A (en) * | 2019-03-08 | 2019-09-20 | 腾讯科技（深圳）有限公司 | Model training method, device, computer readable storage medium and computer equipment |
Non-Patent Citations (1)
| Title |
|---|
| 结合词性特征与卷积神经网络的文本情感分析;何鸿业 等;计算机工程(第11期);第215-220页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111061870A (en) | 2020-04-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20180096219A1 (en) | Neural network combined image and text evaluator and classifier | |
| CN108255857B (en) | Statement detection method and device | |
| US9208441B2 (en) | Information processing apparatus, information processing method, and program | |
| CN106940705A (en) | A kind of method and apparatus for being used to build user's portrait | |
| CN108021651B (en) | Network public opinion risk assessment method and device | |
| CN110533018B (en) | Image classification method and device | |
| CN112487149B (en) | Text auditing method, model, equipment and storage medium | |
| CN110427487B (en) | Data labeling method and device and storage medium | |
| CN111061870B (en) | Article quality evaluation method and device | |
| CN110245227B (en) | Training method and device for text classification fusion classifier | |
| CN111767957B (en) | Log abnormality detection method and device, storage medium and electronic equipment | |
| CN108090041B (en) | Method and device for generating advertisement creativity | |
| CN111506504A (en) | Software development process measurement-based software security defect prediction method and device | |
| CN110490304B (en) | Data processing method and device | |
| CN110968776A (en) | Policy knowledge recommendation method, device storage medium and processor | |
| CN115718831A (en) | Social media robot detection method based on dynamic heterogeneous graph and related device | |
| CN113590945A (en) | Book recommendation method and device based on user borrowing behavior-interest prediction | |
| CN111199728A (en) | Training data acquisition method and device, intelligent sound box and intelligent television | |
| CN115297042A (en) | Method for detecting consistency of web pages under different networks and related equipment | |
| CN111475641B (en) | Data extraction method and device, storage medium and equipment | |
| CN114493853A (en) | Credit rating evaluation method, credit rating evaluation device, electronic device and storage medium | |
| CN113139382A (en) | Named entity identification method and device | |
| CN109598393A (en) | A kind of analysis method and device of the influence information that event generates enterprise | |
| CN110969011B (en) | Text emotion analysis method and device, storage medium and processor | |
| Mandhala et al. | A Novel Study on Tools and Frameworks for Mitigating Bias in Multimodal Datasets |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20200728 Address after: 518000 Nanshan District science and technology zone, Guangdong, Zhejiang Province, science and technology in the Tencent Building on the 1st floor of the 35 layer Applicant after: TENCENT TECHNOLOGY (SHENZHEN) Co.,Ltd. Address before: 100029, Beijing, Chaoyang District new East Street, building No. 2, -3 to 25, 101, 8, 804 rooms Applicant before: Tricorn (Beijing) Technology Co.,Ltd. |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |