Article quality scoring method, article quality scoring device and storage medium
By calculating multiple static feature scores and weights of the article, the total static feature scores are formed, which solves the problem of low accuracy of article quality scores in the prior art, and improves the quality and user experience of article recommendations.
Patent Information
- Application Number
- CN202010873079.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-08-26
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2040-08-26
AI Technical Summary
The prior art fails to fully utilize the static characteristics of the article in the article quality rating, resulting in low rating accuracy, and methods based on user browsing behavior may lead to low-quality articles being recommended to users, affecting the user experience.
By determining multiple static characteristics of the article to be scored, including text length, paragraph, text picture number, text picture clarity, low-quality features and author, static feature scores and weights are calculated to form a total static feature score, and based on this Perform article quality scoring, set feature threshold intervals and additional scoring criteria to improve scoring accuracy.
It achieves efficient and accurate article quality ratings, reduces the possibility of low-quality articles being recommended to users, and improves user experience and satisfaction.
Smart Images

Figure CN112069802B_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the technical field of article quality scoring, and in particular to an article quality scoring method, an article quality scoring device and a storage medium. Background Art
[0002] With the rapid development of Internet technology, accessing the Internet through browsers or other browser-like applications has become an important means for users to obtain information.
[0003] In order to provide users with a good user experience, browsers or browser-like applications often recommend daily new articles to users. However, since there are many low-quality articles among the new articles, if such articles are recommended to users in large quantities, it will cause user loss. In order to reduce user loss and increase user experience, it is particularly important to recommend high-quality articles to users. Therefore, in the related art, articles are rated for quality and recommended to users based on the rating results.
[0004] At present, in the related technology, the quality rating of the articles to be rated is realized by using the text rating features of the articles to be rated and the pre-built rating model. However, since this method only uses the text data of the articles, and does not fully consider other features of the articles, such as the text pictures of the articles, the classification of the articles, etc., the accuracy of the quality rating of the articles to be rated is low.
[0005] In the related art, the browsing behavior information of the user when browsing the target article (also called the article to be rated) is obtained, and the browsing behavior rating of the user on the target article is obtained according to the browsing behavior information and the corresponding browsing behavior coefficient. Finally, the article quality rating of the target article is obtained according to the browsing behavior ratings of the target article obtained by multiple users. However, since this method only uses the browsing behavior data of the user on the article, and does not make full use of the characteristics of the article itself, it will lead to the accuracy of the quality rating of the target article. Furthermore, based on this method, if the quality of the target article is to be evaluated, it is necessary to recommend the target article to multiple users first. During the application process, it is easy to recommend low-quality target articles to users, which in turn brings a bad experience to users.
[0006] Therefore, how to efficiently and accurately score the quality of articles has become a focus of current attention. Summary of the invention
[0007] In order to overcome the problems existing in the related art, the present disclosure provides an article quality scoring method, an article quality scoring device and a storage medium.
[0008] According to a first aspect of an embodiment of the present disclosure, a method for scoring article quality is provided, comprising: determining a plurality of static features contained in an article to be scored, and determining a static feature score and a static feature weight of each of the plurality of static features; multiplying the static feature score corresponding to each static feature in the plurality of static features by the static feature weight and then adding the resultant amounts to obtain a total static feature score; and determining an article quality score of the article to be scored based on the total static feature score.
[0009] In one embodiment, the article quality scoring method also includes: based on the article type and the feature threshold of at least one reference static feature, pre-setting multiple feature threshold intervals, each feature threshold interval corresponds to a weight combination, wherein each weight combination includes the weights of the multiple static features; determining the static feature weight of each static feature in the multiple static features includes: based on the type of the article to be scored, the static feature in the article to be scored corresponding to the reference static feature, and the feature threshold, determining the feature threshold interval corresponding to the static feature in the article to be scored corresponding to the reference static feature; based on the weight combination corresponding to the feature threshold interval, determining the static feature weight of each static feature in the multiple static features.
[0010] In another embodiment, multiple feature threshold intervals are pre-set based on the article type and the feature threshold of at least one reference static feature, including: setting at least one feature threshold for each of the preset multiple reference static features, and setting different feature thresholds for the reference static features of different article types; combining any two feature thresholds among the feature thresholds corresponding to all the reference static features in the multiple reference static features to form multiple feature threshold intervals.
[0011] In another embodiment, the article quality scoring method also includes: determining whether the article to be scored meets preset additional scoring criteria, the additional scoring criteria including bonus criteria and / or subtraction criteria; determining the article quality score of the article to be scored based on the static feature total score, including: if the article to be scored meets the preset additional scoring criteria, then adding and / or subtracting points from the static feature total score according to the bonus criteria and / or subtraction criteria to obtain the article quality score of the article to be scored.
[0012] In yet another embodiment, before determining a plurality of static features included in the article to be scored, the article quality scoring method further includes: determining that the article to be scored is not a low-quality article.
[0013] In yet another embodiment, the article quality scoring method further includes: in response to determining that the article to be scored is a low-quality article, determining that the article quality score of the article to be scored is a minimum score.
[0014] In another embodiment, the static features include one or more of text length, paragraphs, number of text pictures, clarity of text pictures, low-quality features and authors, and determining the static feature score of each of the multiple static features includes: determining the article text length score according to the article text length; determining the paragraph score according to the number of paragraphs; determining the number of text pictures score according to the number of text pictures; determining the text picture clarity score according to the text picture clarity; determining the low-quality feature score according to the low-quality features; and determining the author score according to the author's level.
[0015] According to a second aspect of an embodiment of the present disclosure, there is provided an article quality scoring device, comprising: a static feature determination module, used to determine a plurality of static features contained in an article to be scored, and to determine a static feature score and a static feature weight of each of the plurality of static features; a processing module, used to multiply the static feature score corresponding to each static feature of the plurality of static features by the static feature weight and then add the resultant sums to obtain a total static feature score; and an article quality scoring module, used to determine an article quality score of the article to be scored based on the total static feature score.
[0016] In one embodiment, the article quality scoring device also includes: a feature threshold interval setting module, which is used to pre-set multiple feature threshold intervals based on the article type and the feature threshold of at least one reference static feature, each feature threshold interval corresponds to a weight combination, wherein each weight combination includes the weights of the multiple static features; the static feature determination module is used to: determine the feature threshold interval corresponding to the static feature corresponding to the reference static feature in the article to be scored based on the type of the article to be scored, the static feature in the article to be scored corresponding to the reference static feature, and the feature threshold; determine the static feature weight of each static feature in the multiple static features based on the weight combination corresponding to the feature threshold interval.
[0017] In another embodiment, the feature threshold interval setting module is used to: set at least one feature threshold for each reference static feature in a plurality of preset reference static features, and set different feature thresholds for reference static features of different article types; combine any two feature thresholds among the feature thresholds corresponding to all reference static features in the plurality of reference static features to form a plurality of feature threshold intervals.
[0018] In another embodiment, the article quality scoring device also includes: a judgment module, used to determine whether the article to be scored meets the preset additional scoring criteria, and the additional scoring criteria include a bonus standard and / or a subtraction standard; the article quality scoring module determines the article quality score of the article to be scored based on the static feature total score in the following manner: if the article to be scored meets the preset additional scoring criteria, the static feature total score is added and / or subtracted according to the bonus standard and / or subtraction standard to obtain the article quality score of the article to be scored.
[0019] In yet another embodiment, the article quality scoring device further includes: a module for determining non-low-quality articles, which is used to determine that the article to be scored is a non-low-quality article.
[0020] In another embodiment, the article quality scoring device further includes: a low-quality article processing module for determining that the article to be scored is a low-quality article and determining that the article quality score of the article to be scored is the lowest score.
[0021] In another embodiment, the static features include one or more of text length, paragraphs, number of text pictures, clarity of text pictures, low-quality features and authors, and the static feature determination module determines the static feature score of each of the multiple static features in the following manner: determining the article text length score according to the article text length; determining the paragraph score according to the number of paragraphs; determining the number of text pictures score according to the number of text pictures; determining the text picture clarity score according to the text picture clarity; determining the low-quality feature score according to the low-quality features; and determining the author score according to the author's level.
[0022] The technical solution provided by the embodiments of the present disclosure may include the following beneficial effects: the article quality scoring method provided by the present disclosure makes full use of the various static features of the article to be scored, obtains the total score of the static features by determining the score of each static feature and the weight of the static features, and determines the article quality score of the article to be scored based on the total score of the static features. The article quality score of the article to be scored can be determined efficiently and accurately, and high-quality articles can be recommended to users, laying a foundation for improving user satisfaction in the process of obtaining recommended articles.
[0023] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.
[0025] Figure 1A flow chart showing a current method for quality scoring of articles to be scored.
[0026] Figure 2 A flowchart showing another current method for quality scoring of articles to be scored.
[0027] Figure 3 The figure is a flow chart of a method for scoring article quality according to an exemplary embodiment.
[0028] Figure 4 A schematic diagram showing an application scenario of an application article quality scoring method.
[0029] Figure 5 The figure is a flow chart of another article quality scoring method according to an exemplary embodiment.
[0030] Figure 6 A flow chart for determining a static feature weight for each static feature in a plurality of static features is shown.
[0031] Figure 7 A flow chart of presetting multiple feature threshold intervals based on article type and feature threshold of at least one reference static feature is shown.
[0032] Figure 8 The figure is a flow chart of another article quality scoring method according to an exemplary embodiment.
[0033] Fig. 9 The figure is a flow chart of another article quality scoring method according to an exemplary embodiment.
[0034] Fig.10 The figure is a block diagram of a device for scoring article quality according to an exemplary embodiment.
[0035] Fig.11 The invention is a block diagram of a device for scoring article quality according to an exemplary embodiment. DETAILED DESCRIPTION
[0036] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.
[0037] We are currently in an era of information explosion. Browsers or browser-like applications, such as news push applications, MIUI browsers and other clients, add a large number of articles every day. These new articles are of various types and may include low-quality articles, such as advertising articles, clickbait articles, etc. If such articles are recommended to users in large quantities, the user's reading experience will be greatly reduced, resulting in user loss. In addition, the new articles will also include high-quality articles. If relevant articles are pushed to users based on user characteristics and preferences, the user experience can be improved, thereby better retaining users. Therefore, it is particularly important to efficiently and accurately score the quality of articles.
[0038] Figure 1 A flow chart showing a current method for quality scoring of articles to be scored.
[0039] like Figure 1 As shown in the figure, in the related art, the quality rating of the articles to be rated can be achieved by using the text rating features of the articles to be rated and the pre-built rating model. However, since this method only uses the text data of the articles, and does not fully consider other features of the articles, such as the text pictures of the articles, the classification of the articles, etc., the accuracy of the quality rating of the articles to be rated is low.
[0040] In one example, some articles to be rated have good concept features, chapter structure features, and lexical semantic features. However, if there are no pictures in the text of these articles, it will be boring for users to read, and the visual sense of the articles is also poor, which ultimately leads to a bad user experience. Or, if there are some advertising content in the articles to be rated, it will make users feel disgusted. From a practical point of view, the overall quality of these articles is not high, but under this solution, the quality scores of such articles will be higher.
[0041] Figure 2 A flowchart showing another current method for quality scoring of articles to be scored.
[0042] like Figure 2 As shown, in the related art, it is also possible to obtain the browsing behavior information of the user when browsing the target article (also called the article to be rated), and obtain the browsing behavior score of the user on the target article according to the browsing behavior information and the corresponding browsing behavior coefficient, and finally, obtain the article quality score of the target article according to the browsing behavior scores of multiple users on the target article. However, since this method only uses the browsing behavior data of the user on the article, and does not make full use of the characteristics of the article itself, it will lead to the accuracy of the quality score of the target article.
[0043] Furthermore, under this scheme, if the target article itself is of very good quality, but when collecting the browsing behaviors of multiple users, if most of the selected users are not interested in the article, but in reality there are other users who are not collected and are interested in the article, then according to this scheme, the target article will be unilaterally rated as a low-quality article, reducing the accuracy of the rating.
[0044] Furthermore, in the article quality scoring scheme based on user browsing information, if the target article is to be evaluated for quality, it is necessary to first recommend the article to be rated to multiple users, and then obtain the browsing behavior of multiple users on the rated article. In this process, if the target article recommended to the user is a low-quality article, although the quality score of the article to be rated can be calculated as a low score based on the user's browsing behavior, since this low-quality article has brought a bad experience to the user, it is very likely to cause the loss of the user.
[0045] Therefore, the method of article quality scoring in related technologies needs to be further optimized.
[0046] The disclosed embodiments provide an article quality scoring method, in which scoring is performed based on multiple static features of an article so as to efficiently and accurately determine the article quality score of the article to be scored, recommend high-quality articles to users, and lay a foundation for improving the user's satisfaction in the process of obtaining recommended articles.
[0047] In one embodiment of the present disclosure, article quality scoring is performed before pushing articles to users, so as to further ensure the quality of the pushed articles.
[0048] Figure 3 The figure is a flow chart of a method for scoring article quality according to an exemplary embodiment.
[0049] In an exemplary embodiment of the present disclosure, Figure 3 As shown, the article scoring method may include steps S11 to S13. Each step will be introduced below.
[0050] In step S11, multiple static features included in the article to be rated are determined, and a static feature score and a static feature weight of each of the multiple static features are determined.
[0051] In an example of the present disclosure, static features can be understood as non-dynamic features of the article to be rated. For example, the static features of the article to be rated may include the length of the text, paragraphs, the number of text images, the clarity of the text images, low-quality features, and the author. In the present disclosure, no specific limitation is made to static features. If new static features are subsequently used to evaluate the quality of the article, the new static features can still be applied to the article quality scoring method involved in the present disclosure.
[0052] From the above description, it can be seen that the static features of the article to be evaluated basically cover all aspects of the article's own features, such as the text, pictures, paragraphs, authors, low-quality features, etc. In the present disclosure, the article quality is scored based on multiple static features of the article to be scored, which can improve the accuracy of the scoring.
[0053] The disclosed embodiment can determine static feature scores and static feature weights for multiple static features of the article to be scored, respectively, and perform article quality scoring based on the static feature scores and static feature weights of the multiple static features of the article to be scored, so as to more accurately perform quality scoring on the article to be scored.
[0054] In step S12, the static feature score corresponding to each static feature among the multiple static features is multiplied by the static feature weight and then added up to obtain a total static feature score.
[0055] During the application process, the static feature score corresponding to each of the multiple static features may be multiplied by the static feature weight and then added up to obtain the total static feature score of the article to be rated.
[0056] It should be noted that each static feature weight can be determined according to the article type, the article body and pictures of the article to be evaluated. Each static feature weight can also be determined according to other methods. In this embodiment, no specific limitation is made to each static feature weight.
[0057] In step S13, the article quality score of the article to be scored is determined based on the total score of the static features.
[0058] In one example, the article quality score of the article to be rated can be determined according to the total static feature score determined by each static feature. In a possible implementation, the article to be rated can be divided into high-quality articles, medium-quality articles and low-quality articles based on the article quality scores of the article to be rated.
[0059] In one possible implementation, when a browser or a browser-like application adds a new article that needs to be pushed to a user, the new article can be scored for quality based on the article quality scoring method provided in the embodiment of the present disclosure, and based on the result of the article quality scoring, the new article can be recommended to the user in a targeted manner.
[0060] In one example, if some newly added articles have low quality scores and are evaluated as low-quality articles, low-quality articles can be filtered out when recommending articles to users as needed. For example, through user behavior analysis, it is found that a certain user is very disgusted with clickbait or advertising articles. Then, when recommending articles to the user, low-quality articles belonging to clickbait or advertising can be actively filtered out to prevent the user from receiving such articles.
[0061] Figure 4 A schematic diagram showing an application scenario of an application article quality scoring method. Figure 4 1 is an article, 2 is an article quality scoring method involved in the present disclosure, 3 is a personalized recommendation system, and 4 is a user. When a new article is added, the quality score of the article is determined according to the article quality method involved in the present disclosure, and then the personalized recommendation system recommends articles of different quality scores to the user according to the user's characteristics. When the article is exposed to the user, the user will give feedback on the article, and then the user's behavior data is collected based on the user's feedback, so as to recommend articles to the user more accurately in the future.
[0062] In an example, Figure 4 As shown, if user 4 is very interested in food article 1, and through the analysis of the user's previous behavior data, it is found that the user has a short reading time for food article 1 with a low article quality score and has no other operations, but has a long reading time for some articles with high article quality scores and often comments on or collects such articles. Then, for the newly added food article 1 in the browser or a browser-like application, the article quality scoring method 2 provided in the embodiment of the present disclosure can be used to score the newly added food article 1, and the food article with a high article quality score can be recommended to user 4 through the personalized recommendation system 3.
[0063] In another example, a user often reads or comments on high-quality articles and medium-quality articles and does not show obvious preference for article types. In this case, high-quality articles and medium-quality articles can be selected by multiple categories to recommend articles to the user.
[0064] It should be noted that after an article is recommended to a user, the user may give feedback on the article. In one embodiment, the user's behavior data can also be obtained based on the feedback given by the user to lay a foundation for recommending articles that better meet the user's requirements to the user in the future.
[0065] The article quality scoring method provided by the present disclosure makes full use of the various static features of the article to be scored, obtains the total score of the static features by determining the score of each static feature and the weight of the static feature, and determines the article quality score of the article to be scored based on the total score of the static features. The present disclosure can efficiently and accurately determine the article quality score of the article to be scored, recommend high-quality articles to users, and lay a foundation for improving the user's satisfaction in the process of obtaining recommended articles.
[0066] The present disclosure will illustrate the process of article quality scoring through the following embodiments.
[0067] In the disclosed embodiment, the static feature score determination process is first described.
[0068] In an exemplary embodiment of the present disclosure, the static features include one or more of text length, paragraphs, number of text images, clarity of text images, low-quality features, and authors. The static feature score of each of the multiple static features may be determined in the following manner.
[0069] In one example, the article body length score can be determined based on the article body length. The article body length score can be determined by the following formula:
[0070]
[0071] Among them, bodyLen represents the length of the body of the article to be rated. l1 and l2 represent two thresholds of the length of the body of the article to be rated corresponding to different article types, among which l1 is a smaller threshold (for example, X in the above text). 小 =600), l2 is a larger threshold (e.g., X 大 =1000). w1 and w2 represent two parameters, wherein w1 and w2 can be set according to l1 and l2.
[0072] In one example, the paragraph score can be determined according to the number of paragraphs. The paragraph score can be determined by the following formula:
[0073]
[0074] Among them, paragraphNum represents the number of paragraphs in the article.
[0075] In one example, the score of the number of body pictures can be determined according to the number of body pictures. The score of the number of body pictures can be determined by the following formula:
[0076]
[0077] Where imgNum represents the number of text images. n2 represents the larger threshold of the two thresholds of the number of text images of different article types in the article to be rated (for example, P 大 =5).
[0078] In one example, the clarity score of the main text image can be determined according to the clarity of the main text image. The clarity score of the main text image can be determined by the following formula:
[0079]
[0080] Among them, imgClarity represents the clarity value of the image, and n represents the number of text images.
[0081] In one example, a low-quality feature score can be determined based on the low-quality feature. The low-quality feature score can be determined by the following formula:
[0082] negFeatureScore=min(negFeatureScore1, negFeatureScore2...)
[0083] Among them, the total score of low-quality features is the minimum value of all low-quality feature scores.
[0084] In one example, the author score can be determined according to the author's level. The author score can be determined by the following formula:
[0085] authorScore=0.2*level
[0086] Where level represents the author level. In one example, the author level can be divided into 5 levels: 1, 2, 3, 4, 5. The author score is defined according to the author level. The higher the author level, the higher the author score.
[0087] Furthermore, the static feature scores of the above six static features can be multiplied by the static feature weights and then added up to obtain the total static feature score. The total static feature score can be determined by the following formula:
[0088]
[0089] Wherein: i represents the ith static feature, factor represents the static feature score, and weight represents the static feature weight corresponding to the static feature. In the disclosed embodiment, a total of the above 6 static features are involved: text length, paragraphs, number of text images, text image clarity, low-quality features, and author.
[0090] The following is an explanation of the process of determining the static feature weights in the embodiments of the present disclosure.
[0091] In one implementation of the disclosed embodiment, the weights of static features can be pre-set, and the weights of the static features corresponding to each article type can be formed into a weight combination. Each weight combination includes the weights of multiple static features. When determining the weights of static features, the weights of static features can be determined according to the preset weight combination, and the article quality can be scored.
[0092] Figure 5 The figure is a flow chart of another article quality scoring method according to an exemplary embodiment.
[0093] In an exemplary embodiment of the present disclosure, Figure 5 As shown, the article quality scoring method may include steps S21 to S24. Each step will be introduced below.
[0094] In step S21, based on the article type and the feature threshold of at least one reference static feature, multiple feature threshold intervals are pre-set, each feature threshold interval corresponds to a weight combination, wherein each weight combination includes weights of multiple static features.
[0095] In step S22, multiple static features included in the article to be scored are determined, and a static feature score and a static feature weight of each of the multiple static features are determined.
[0096] In step S23, the static feature score corresponding to each static feature in the multiple static features is multiplied by the static feature weight and then added up to obtain a total static feature score.
[0097] In step S24, the article quality score of the article to be scored is determined based on the total score of the static features.
[0098] Among them, steps S22 to S24 are respectively the same as steps S11 to S13 in the above embodiment, and their related explanations and descriptions as well as beneficial effects are referred to the above descriptions of steps S11 to S13, which will not be repeated here. Step S21 will be described in detail below.
[0099] During the application process, multiple feature threshold intervals can be pre-set based on the article type and the feature threshold of at least one reference static feature. Among them, the article type can include food articles, current affairs articles, comic articles, etc. The reference static feature can be understood as a feature that can reflect the characteristics of the article to a certain extent. In one example, the reference static feature can be the length of the article body, the number of text pictures, etc. In the present disclosure, no specific limitation is made to the reference static feature.
[0100] In a possible embodiment, each reference static feature may have at least one feature threshold. In one example, for a food article, when the reference static feature is the length of the text, there may be two corresponding feature thresholds, for example, a larger threshold (1000) and a smaller threshold (800). Further, based on the larger threshold and the smaller threshold, three feature threshold intervals may be set for the static feature regarding the length of the text (L≥1000, 800<L<1000, L≤800, where L represents the length of the text).
[0101] It should be noted that each feature threshold interval corresponds to a weight combination, and each weight combination includes weights of multiple static features. In the application process, the static feature weights of the static features in the article to be rated can be determined based on the weights of the static features corresponding to the feature threshold interval.
[0102] Furthermore, the static feature score of the article to be rated is multiplied by the static feature weight and then added up to obtain a total static feature score, and the article quality score of the article to be rated is determined based on the total static feature score.
[0103] According to the above description, in the embodiment of the present disclosure, the static feature weight of the static feature in the article to be scored can be determined based on the weight of the corresponding static feature within the feature threshold range.
[0104] The present disclosure will illustrate the process of determining the static feature weight of each static feature in a plurality of static features in a to-be-rated article based on the weight of the static feature corresponding to the feature threshold interval through the following embodiments.
[0105] Figure 6 A flow chart for determining a static feature weight for each static feature in a plurality of static features is shown.
[0106] In an exemplary embodiment of the present disclosure, determining the static feature weight of each static feature in the plurality of static features may include step S31 and step S32. Each step will be described below.
[0107] In step S31, based on the type of the article to be rated, the static feature corresponding to the reference static feature in the article to be rated, and the feature threshold, a feature threshold interval corresponding to the reference static feature in the article to be rated is determined.
[0108] During the application process, due to the different types of articles, for the same reference static feature, the feature thresholds of the static feature are also different. In one example, for food articles, when the reference static feature is the length of the text, there may be two corresponding feature thresholds, for example, a larger threshold (1000) and a smaller threshold (800). For political articles, when the reference static feature is the length of the text, there may be two corresponding feature thresholds, for example, a larger threshold (2000) and a smaller threshold (1000).
[0109] In step S32, based on the weight combination corresponding to the feature threshold interval, the static feature weight of each static feature in the plurality of static features is determined.
[0110] It should be noted that, in addition to the weight combination, the feature threshold interval can also represent the value range of the reference static feature. By judging whether the static feature corresponding to the reference static feature in the article to be rated is within the feature threshold interval, it is determined whether the static feature weight of the article to be rated can be determined according to the weight combination included in the feature threshold interval.
[0111] The present disclosure will illustrate the process of determining multiple feature threshold intervals through the following embodiments.
[0112] Figure 7 A flow chart of presetting multiple feature threshold intervals based on article type and feature threshold of at least one reference static feature is shown.
[0113] In an exemplary embodiment of the present disclosure, presetting a plurality of feature threshold intervals based on the article type and the feature threshold of at least one reference static feature may include step S41 and step S42.
[0114] In step S41, at least one feature threshold is set for each of the preset multiple reference static features, and different feature thresholds are set for reference static features of different article types.
[0115] In step S42, any two feature thresholds among the feature thresholds corresponding to all the reference static features in the multiple reference static features are combined to form a plurality of feature threshold intervals.
[0116] In one example, two reference static features (e.g., text length and the number of text images) can be determined. Now, taking food articles as an example, the text length can include two thresholds, and the number of text images can also include two thresholds. For ease of description, let the larger threshold of text length be X. 大 (1000); the smaller threshold of the text length is X 小 (800); The larger threshold of the number of text images is Y 大(5) The smaller threshold of the number of text images is Y 小 (0). Further, based on X 大 and X 小 Three feature threshold intervals can be set for the static feature about the length of the text (L≥1000, 800<L<1000, L≤800, where L represents the length of the text); based on Y 大 and Y 小 Three feature threshold intervals (P≥5, 0<P<5, P≤5, where P represents the number of text pictures) can be set for the static feature about the number of text pictures.
[0117] Furthermore, the two reference static features of text length and text image number can be combined to combine the feature threshold intervals to form 9 (3*3) different feature threshold intervals, forming a feature threshold interval similar to a nine-grid grid. Please refer to Table 1.
[0118] Table 1 Nine-grid grid with different feature threshold intervals
[0119] <![CDATA[L≤X 小 ,P≥Y 大 ]]> <![CDATA[X 小 <L<X 大 ,P≥Y 大 ]]> <![CDATA[L≥X 大 ,P≥Y 大 ]]> <![CDATA[L≤X 小 ,AND 小 <P<Y 大 ]]> <![CDATA[X 小 <L<X 大 ,AND 小 <P<Y 大 ]]> <![CDATA[L≥X 大 ,AND 小 <P<Y 大 ]]> <![CDATA[L≤X 小 ,P≤Y 小 ]]> <![CDATA[X 小 <L<X 大 ,P≤Y 小 ]]> <![CDATA[L≥X 大 ,P≤Y 小 ]]>
[0120] It should be noted that each article type can correspond to a grid in the nine-square grid in Table 1. For the static features of the articles to be rated, different grids (feature threshold intervals) have different weight combinations, and articles of different article types to be rated will have the same weight combination in the same grid (feature threshold interval).
[0121] In the application process, for an article to be rated, the corresponding grid in the nine-square grid (L≤X) can be found according to the article type of the article to be rated and the value of the static feature corresponding to the reference static feature (for example, the length of the text is 600 and the number of text images is 0). 小 , P≤Y 小 ), and determine the static feature weight by combining the weights in the grid.
[0122] In one example, for food articles, the two thresholds for the length of the text may be 800 and 1000, and the two thresholds for the number of text images may be 0 and 5. When the length of the text of an article to be rated is 600 and the number of text images is 0, the feature threshold interval can be determined as (L≤X 小 , P≤Y 小 ); When the length of the text of an article to be rated is 900 and the number of text images is 3, the feature threshold interval can be determined as (X 小 <L<X 大 , Y 小 <P<Y 大); When the length of the text of an article to be rated is 950 and the number of text images is 5, the feature threshold interval can be determined as (X 小 <L<X 大 , P ≥ Y 大 ); When the length of the text of an article to be rated is 1200 and the number of text images is 6, the feature threshold interval can be determined as (L≥X 大 , P ≥ Y 大 ).
[0123] Furthermore, the static feature score of the article to be rated can be multiplied by the static feature weight and then added up to obtain the total static feature score, and the article quality score of the article to be rated can be determined based on the total static feature score.
[0124] During the application process, if the article to be rated meets some additional rating criteria, it will affect the final article quality rating of the article to be rated.
[0125] The following embodiments of the present disclosure will illustrate the process of scoring the quality of an article when the article to be scored meets some additional scoring criteria.
[0126] Figure 8 The figure is a flow chart of another article quality scoring method according to an exemplary embodiment.
[0127] In an exemplary embodiment of the present disclosure, Figure 8 As shown, the article quality scoring method includes steps S51 to S54. Each step will be introduced below.
[0128] In step S51, multiple static features included in the article to be scored are determined, and a static feature score and a static feature weight of each of the multiple static features are determined.
[0129] In step S52, the static feature score corresponding to each static feature among the multiple static features is multiplied by the static feature weight and then added up to obtain a total static feature score.
[0130] In step S53, it is determined whether the article to be rated meets the preset additional rating criteria, wherein the additional rating criteria include a score increase criterion and / or a score reduction criterion.
[0131] In step S54, if the article to be rated meets the preset additional rating criteria, the total score of the static features is added and / or subtracted according to the score addition criteria and / or score subtraction criteria to obtain the article quality score of the article to be rated.
[0132] Among them, step S51-step S52 are respectively the same as step S11-step S12 in the above embodiment, and the relevant explanations and descriptions and beneficial effects thereof are referred to the above descriptions of step S11-step S12, which will not be repeated here. Step S53 and step S54 will be described in detail below.
[0133] During the application process, it can be determined whether the article to be rated meets the preset attachment scoring criteria. Among them, the additional scoring criteria include bonus criteria and / or deduction criteria. In one example, when it is detected that the punctuation mark at the end of the article to be evaluated is ":", it means that the content of the article to be rated has not ended and its content is incomplete. Therefore, it meets the preset deduction criteria. During the application process, the total score of static features can be deducted to obtain the article quality score of the article to be rated. In another example, since there are subheadings for paragraphs in the body of the article to be rated, the preset bonus criteria are met. During the application process, the total score of static features can be added to obtain the article quality score of the article to be rated.
[0134] It should be noted that the preset additional scoring criteria can be adjusted according to actual conditions, and in the present disclosure, no specific limitation is made to the preset additional scoring criteria.
[0135] Since it is easy to judge whether the article is a low-quality article (clickbait, advertisement, competitor, feudal superstition, low style, repeated content, etc.) based on the title and text of the article to be rated, when obtaining an article to be rated, it can be judged whether the article is a low-quality article first.
[0136] The present disclosure will illustrate the article quality scoring method through the following examples.
[0137] In an exemplary embodiment of the present disclosure, before determining the multiple static features included in the article to be rated, it is also possible to first determine whether the article to be rated is a non-low-quality article. If it is determined that the article to be rated is not a non-low-quality article, the article can be directly defined as a low-quality article without performing steps such as determining the multiple static features included in the article to be rated. Through this embodiment, the calculation process of determining the article quality score of the article to be rated can be reduced.
[0138] In an exemplary embodiment of the present disclosure, in response to determining that the article to be rated is a low-quality article, the article quality score of the article to be rated is determined to be the lowest score. In one example, the lowest score can be 0.
[0139] The following disclosed embodiments illustrate the article quality scoring method involved in the above embodiments in combination with practical applications.
[0140] Fig. 9 The figure is a flow chart of another article quality scoring method according to an exemplary embodiment.
[0141] like Fig. 9 As shown, in one example, the title, text, text pictures, article classification, and author information of the article to be rated can be obtained. Based on the above information of the article to be rated, it is determined whether the article to be rated is a low-quality article or has repeated content. If it is determined that the article to be rated is a low-quality article or has repeated content, the article quality score of the article is directly set to 0, and the following steps are not performed.
[0142] If it is determined that the article to be rated is not a low-quality article and there is no content duplication, the static feature scores of the article to be rated are further calculated (for example, text length, paragraphs, number of text pictures, text picture clarity, low-quality features and author). And according to the type of article, text length and number of text pictures, the weight combination corresponding to each grid (feature threshold interval) in the nine-square grid of Table 1 is searched to determine the static feature weights corresponding to each static feature of the article to be rated. And each static feature is multiplied by the static feature weight and then added to obtain the total quality score of the static features of the article to be rated.
[0143] Further, it is determined whether the article to be rated meets the preset additional rating criteria. The additional rating criteria include a score-adding criterion and / or a score-reducing criterion. If the article to be rated meets the preset additional rating criteria, the static feature total score is added and / or subtracted according to the score-adding criterion and / or the score-reducing criterion to obtain the final article quality score of the article to be rated.
[0144] As can be seen from the above description, the article quality scoring method provided by the present disclosure makes full use of the various static features of the article to be scored, obtains the total score of the static features by determining the score of each static feature and the weight of the static feature, and determines the article quality score of the article to be scored based on the total score of the static features. The present disclosure can efficiently and accurately determine the article quality score of the article to be scored, recommend high-quality articles to users, and lay a foundation for improving the user's satisfaction in the process of obtaining recommended articles.
[0145] The article quality scoring method provided by the embodiment of the present disclosure aims at the shortcoming of low accuracy of the quality scoring scheme based solely on text involved in the related technical scheme. It makes full use of the various static features of the article itself, and obtains the scores of various static features of the article based on the title, body text, body pictures, author and other information of the article. Then, different weights are assigned to each static feature according to the characteristics of the body length and the number of body pictures of articles of different categories, so as to accurately identify whether the article is a low-quality article or a high-quality article. If the article is a clickbait, an advertisement, a competing product, feudal superstition, or a low-style article, the article has repeated content, the article is short and has no pictures, the article content is incomplete, and the article content is water text, etc., it is defined as a low-quality article. If the article content is deep, professional, rich in pictures and texts, well-structured, and complete in structure, it is defined as a high-quality article, and others can be defined as medium-quality articles. This technical solution greatly improves the accuracy of article quality scoring.
[0146] The article quality scoring method provided by the embodiment of the present disclosure aims to address the shortcoming of related technical solutions that determining article quality scores based on user browsing behavior can easily cause user loss. The quality of the article is evaluated before the article is recommended to the user, thereby avoiding recommending some low-quality articles to the user, and different articles can be recommended to the user according to different user characteristics, thereby bringing better performance to the user, satisfying his interests and hobbies, and retaining the user.
[0147] Based on the same concept, the embodiment of the present disclosure also provides an article quality scoring device.
[0148] It is understandable that in order to achieve the above functions, the article quality scoring device provided by the embodiment of the present disclosure includes hardware structures and / or software modules corresponding to the execution of each function. In combination with the units and algorithm steps of each example disclosed in the embodiment of the present disclosure, the embodiment of the present disclosure can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the technical solution of the embodiment of the present disclosure.
[0149] Fig.10 is a block diagram of an article quality scoring device according to an exemplary embodiment. Fig.10 It can be seen that the article quality scoring device includes a static feature determination module 110, a processing module 120 and an article quality scoring module 130. Each module will be introduced below.
[0150] The static feature determination module 110 may be configured to: determine a plurality of static features included in the article to be rated, and determine a static feature score and a static feature weight of each of the plurality of static features.
[0151] The processing module 120 may be configured to: multiply the static feature score corresponding to each static feature in the plurality of static features by the static feature weight and then add the resultant sums to obtain a total static feature score.
[0152] The article quality scoring module 130 may be configured to determine the article quality score of the article to be scored based on the static feature total score.
[0153] In an exemplary embodiment of the present disclosure, the article quality scoring device further includes a module for setting a feature threshold interval.
[0154] The feature threshold interval setting module may be configured to: pre-set a plurality of feature threshold intervals based on the article type and the feature threshold of at least one reference static feature, each feature threshold interval corresponding to a weight combination, wherein each weight combination includes the weights of the plurality of static features;
[0155] The static feature determination module 110 can be configured to: determine a feature threshold interval corresponding to a static feature corresponding to a reference static feature in the article to be scored based on the type of the article to be scored, the static feature corresponding to the reference static feature in the article to be scored, and the feature threshold; determine a static feature weight of each static feature in multiple static features based on a weight combination corresponding to the feature threshold interval.
[0156] In an exemplary embodiment of the present disclosure, the feature threshold interval setting module can be configured to: set at least one feature threshold for each reference static feature in a plurality of preset reference static features, and set different feature thresholds for reference static features of different article types; combine any two feature thresholds among the feature thresholds corresponding to all reference static features in the plurality of reference static features to form a plurality of feature threshold intervals.
[0157] In an exemplary embodiment of the present disclosure, the article quality scoring device further includes a judgment module.
[0158] The judgment module may be configured to determine whether the article to be rated meets a preset additional rating standard, where the additional rating standard includes a score increase standard and / or a score reduction standard.
[0159] The article quality scoring module 130 can determine the article quality score of the article to be scored based on the static feature total score in the following manner: if the article to be scored meets the preset additional scoring criteria, the static feature total score is added and / or subtracted according to the addition and / or subtraction criteria to obtain the article quality score of the article to be scored.
[0160] In an exemplary embodiment of the present disclosure, the article quality scoring device further includes a module for determining non-low-quality articles.
[0161] The module for determining non-low-quality articles may be configured to: determine that the article to be rated is a non-low-quality article.
[0162] In an exemplary embodiment of the present disclosure, the article quality scoring device further includes a module for processing low-quality articles.
[0163] The low-quality article processing module may be configured to: in response to determining that the article to be rated is a low-quality article, determine the article quality score of the article to be rated to be the lowest score.
[0164] In an exemplary embodiment of the present disclosure, static features include one or more of text length, paragraphs, number of text pictures, clarity of text pictures, low-quality features, and authors. The static feature determination module 110 can determine the static feature score of each of the multiple static features in the following manner: determine the article body length score according to the article body length; determine the paragraph score according to the number of paragraphs; determine the number of text pictures score according to the number of text pictures; determine the text picture clarity score according to the text picture clarity; determine the low-quality feature score according to the low-quality features; and determine the author score according to the author's level.
[0165] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.
[0166] Fig.11 2 is a block diagram of an apparatus 200 for scoring article quality according to an exemplary embodiment. For example, the apparatus 200 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, etc.
[0167] Reference Fig.11 , the device 200 may include one or more of the following components: a processing component 202 , a memory 204 , a power component 206 , a multimedia component 208 , an audio component 210 , an input / output (I / O) interface 212 , a sensor component 214 , and a communication component 216 .
[0168] The processing component 202 generally controls the overall operation of the device 200, such as operations associated with display, phone calls, data communications, camera operations, and recording operations. The processing component 202 may include one or more processors 220 to execute instructions to perform all or part of the steps of the above-described method. In addition, the processing component 202 may include one or more modules to facilitate interaction between the processing component 202 and other components. For example, the processing component 202 may include a multimedia module to facilitate interaction between the multimedia component 208 and the processing component 202.
[0169] The memory 204 is configured to store various types of data to support operations on the device 200. Examples of such data include instructions for any application or method operating on the device 200, contact data, phone book data, messages, pictures, videos, etc. The memory 204 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.
[0170] The power component 206 provides power to the various components of the device 200. The power component 206 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the device 200.
[0171] The multimedia component 208 includes a screen that provides an output interface between the device 200 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 208 includes a front camera and / or a rear camera. When the device 200 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera may receive external multimedia data. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.
[0172] The audio component 210 is configured to output and / or input audio signals. For example, the audio component 210 includes a microphone (MIC), and when the device 200 is in an operation mode, such as a call mode, a recording mode, and a speech recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in the memory 204 or sent via the communication component 216. In some embodiments, the audio component 210 also includes a speaker for outputting audio signals.
[0173] I / O interface 212 provides an interface between processing component 202 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include but are not limited to: a home button, a volume button, a start button, and a lock button.
[0174] The sensor assembly 214 includes one or more sensors for providing various aspects of the status assessment of the device 200. For example, the sensor assembly 214 can detect the open / closed state of the device 200, the relative positioning of components, such as the display and keypad of the device 200, the sensor assembly 214 can also detect the position change of the device 200 or a component of the device 200, the presence or absence of user contact with the device 200, the orientation or acceleration / deceleration of the device 200 and the temperature change of the device 200. The sensor assembly 214 can include a proximity sensor configured to detect the presence of a nearby object without any physical contact. The sensor assembly 214 can also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 214 can also include an accelerometer, a gyroscope sensor, a magnetic sensor, a pressure sensor or a temperature sensor.
[0175] The communication component 216 is configured to facilitate wired or wireless communication between the device 200 and other devices. The device 200 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 216 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 216 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.
[0176] In an exemplary embodiment, the apparatus 200 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above-mentioned article quality scoring method.
[0177] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 204 including instructions, and the instructions can be executed by the processor 220 of the device 200 to complete the above-mentioned article quality scoring method. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.
[0178] It is to be understood that in the present disclosure, "plurality" refers to two or more than two, and other quantifiers are similar. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the associated objects before and after are in an "or" relationship. The singular forms "a", "the" and "the" are also intended to include plural forms, unless the context clearly indicates other meanings.
[0179] It is further understood that the terms "first", "second", etc. are used to describe various information, but such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other, and do not indicate a specific order or degree of importance. In fact, the expressions "first", "second", etc. can be used interchangeably. For example, without departing from the scope of the present disclosure, the first information can also be referred to as the second information, and similarly, the second information can also be referred to as the first information.
[0180] It will be further understood that the terms “center”, “longitudinal”, “lateral”, “front”, “back”, “up”, “down”, “left”, “right”, “vertical”, “horizontal”, “top”, “bottom”, “inside”, “outside”, etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the present embodiment and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation.
[0181] In the accompanying drawings, the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The described embodiments are part of the embodiments of the present disclosure, rather than all of the embodiments. The embodiments described above with reference to the accompanying drawings are exemplary and are intended to be used to explain the present disclosure, and should not be construed as limitations on the present disclosure. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present disclosure. The embodiments of the present disclosure are described in detail above in conjunction with the accompanying drawings.
[0182] It can be further understood that, unless otherwise specified, “connection” includes a direct connection without other components between the two, and also includes an indirect connection with other components between the two.
[0183] It is further understood that, although the operations are described in a specific order in the drawings in the embodiments of the present disclosure, it should not be understood as requiring the operations to be performed in the specific order shown or in a serial order, or requiring the execution of all the operations shown to obtain the desired results. In certain environments, multitasking and parallel processing may be advantageous.
[0184] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.
[0185] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.
Claims
1. A method for scoring article quality, characterized in that: include: Determine a plurality of static features included in the article to be scored, and determine a static feature score and a static feature weight of each of the plurality of static features; The static feature score corresponding to each static feature in the multiple static features is multiplied by the static feature weight and then added up to obtain a total static feature score; Determining the article quality score of the article to be scored based on the total score of the static features; The method further comprises: Based on the article type and the feature threshold of at least one reference static feature, a plurality of feature threshold intervals are pre-set, each feature threshold interval corresponds to a weight combination, wherein each weight combination includes the weights of the plurality of static features; The determining of the static feature weight of each static feature in the plurality of static features comprises: Determine a feature threshold interval corresponding to the static feature in the article to be rated that corresponds to the reference static feature based on the type of the article to be rated, the static feature in the article to be rated that corresponds to the reference static feature, and the feature threshold; Determining a static feature weight of each static feature of the plurality of static features based on a weight combination corresponding to the feature threshold interval; The static features include text length and paragraphs, wherein the text length score of the article to be rated is determined according to the text length of the article to be rated, and the paragraph score is determined according to the number of paragraphs; The text length score is determined in the following manner: , Wherein, bodyLen represents the length of the body of the article to be rated, l1 and l2 represent two thresholds of the length of the body of the article to be rated corresponding to different article types, w1 and w2 represent two parameters, and w1 and w2 are set based on l1 and l2; The paragraph score is determined in the following manner: , Wherein, paragraphNum represents the number of paragraphs in the article to be rated.
2. The article quality scoring method according to claim 1, characterized in that: Based on the article type and the feature threshold of at least one reference static feature, multiple feature threshold intervals are pre-set, including: At least one feature threshold is set for each of the preset multiple reference static features, and different feature thresholds are set for the reference static features of different article types; Any two feature thresholds among the feature thresholds corresponding to all the reference static features in the multiple reference static features are combined to form multiple feature threshold intervals.
3. The article quality scoring method according to claim 1, characterized in that: The article quality scoring method also includes: Determine whether the article to be rated meets a preset additional rating standard, wherein the additional rating standard includes a score increase standard and / or a score reduction standard; Determining the article quality score of the article to be scored based on the total score of the static features includes: If the article to be rated meets the preset additional rating criteria, the static feature total score is added and / or subtracted according to the score addition and / or subtraction criteria to obtain the article quality score of the article to be rated.
4. The article quality scoring method according to claim 1, characterized in that: Before determining the multiple static features contained in the article to be scored, the article quality scoring method further includes: Determine that the article to be rated is not a low-quality article.
5. The article quality scoring method according to claim 4, characterized in that: The method further comprises: In response to determining that the article to be rated is a low-quality article, determining the article quality score of the article to be rated to be the lowest score.
6. The article quality scoring method according to claim 1, characterized in that: The static features also include the number of text images, the clarity of the text images, low-quality features, and multiple authors. The determining of the static feature score of each of the multiple static features includes: Determine the score of the number of main text pictures according to the number of main text pictures; Determine the clarity score of the main text image according to the clarity of the main text image; Determining a low-quality feature score based on the low-quality features; and Author scores are determined based on the author's level.
7. An article quality scoring device, characterized in that: include: A static feature determination module is used to determine a plurality of static features contained in the article to be rated, and to determine a static feature score and a static feature weight of each of the plurality of static features; A processing module, configured to multiply the static feature score corresponding to each static feature in the plurality of static features by the static feature weight and then add the resultant to obtain a total static feature score; An article quality scoring module, used to determine the article quality score of the article to be scored based on the total score of the static features; The device also includes: A feature threshold interval setting module is used to pre-set multiple feature threshold intervals based on the article type and the feature threshold of at least one reference static feature, each feature threshold interval corresponds to a weight combination, wherein each weight combination includes the weights of the multiple static features; The static feature determination module is used to: determine the feature threshold interval corresponding to the static feature corresponding to the reference static feature in the article to be rated based on the type of the article to be rated, the static feature corresponding to the reference static feature in the article to be rated, and the feature threshold; determine the static feature weight of each static feature in the multiple static features based on the weight combination corresponding to the feature threshold interval; The static features include text length and paragraphs, wherein the text length score of the article to be rated is determined according to the text length of the article to be rated, and the paragraph score is determined according to the number of paragraphs; The text length score is determined in the following manner: , Wherein, bodyLen represents the length of the body of the article to be rated, l1 and l2 represent two thresholds of the length of the body of the article to be rated corresponding to different article types, w1 and w2 represent two parameters, and w1 and w2 are set based on l1 and l2; The paragraph score is determined in the following manner: , Wherein, paragraphNum represents the number of paragraphs in the article to be rated.
8. The article quality scoring device according to claim 7, characterized in that: The module for setting feature threshold interval is used for: At least one feature threshold is set for each of the preset multiple reference static features, and different feature thresholds are set for the reference static features of different article types; Any two feature thresholds among the feature thresholds corresponding to all the reference static features in the multiple reference static features are combined to form multiple feature threshold intervals.
9. The article quality scoring device according to claim 7, characterized in that: The device also includes: A judgment module, used to determine whether the article to be rated meets a preset additional rating standard, wherein the additional rating standard includes a score increase standard and / or a score reduction standard; The article quality scoring module determines the article quality score of the article to be scored based on the static feature total score in the following manner: If the article to be rated meets the preset additional rating criteria, the static feature total score is added and / or subtracted according to the score addition and / or subtraction criteria to obtain the article quality score of the article to be rated.
10. The article quality scoring device according to claim 7, characterized in that: The device also includes: The module for determining non-low-quality articles is used to determine whether the article to be rated is a non-low-quality article.
11. The article quality scoring device according to claim 10, characterized in that: The device also includes: The low-quality article processing module is used to determine that the article to be rated is a low-quality article and to determine that the article quality score of the article to be rated is the lowest score.
12. The article quality scoring device according to claim 7, characterized in that: The static features also include the number of text images, text image clarity, low-quality features, and multiple authors. The static feature determination module determines the static feature score of each of the multiple static features in the following manner: Determine the score of the number of main text pictures according to the number of main text pictures; Determine the clarity score of the main text image according to the clarity of the main text image; Determining a low-quality feature score based on the low-quality features; and Author scores are determined based on the author's level.
13. An image processing device, characterized in that: include: processor; a memory for storing processor-executable instructions; Wherein, the processor is configured to: execute the article quality scoring method described in any one of claims 1 to 6.
14. A non-transitory computer-readable storage medium, when the instructions in the storage medium are executed by a processor of a mobile terminal, enables the mobile terminal to execute the article quality scoring method described in any one of claims 1 to 6.
