Public opinion information recommendation method and device, storage medium and terminal

Abstract

The embodiment of the invention provides a public opinion information recommendation method and device, a storage medium and a terminal, and the method comprises the steps: obtaining one or more pieces of public opinion information related to a target public opinion theme to form a public opinion information set after the target public opinion theme is determined; obtaining one or more of correlation dimension information, source dimension information and propagation dimension information of each piece of public opinion information in the public opinion information set; and according to the scoring rule corresponding to each piece of dimension information, determining the score of each piece of dimension information, and calculating the value score of each piece of public opinion information in combination with the weight of each piece of dimension information. And performing public opinion information recommendation according to the value score sequence of all public opinion information in the public opinion information set. By applying the method, the public opinion information can be automatically analyzed and screened from multiple dimensions, the efficiency and authority of public opinion information acquisition are improved, and real-time updating is supported, so that enterprises or governments and the like can acquire the latest public opinion information in real time and quickly take countermeasures, and potential business risks are avoided.

Description

Technical Field

[0001] The present invention relates to the field of mobile information screening, and in particular to a method, device, storage medium and terminal for recommending public opinion information. Background Art

[0002] With the rapid development of mobile Internet, Android applications have become an important part of the daily life of smartphone users around the world. However, due to the openness of the Android system, illegal applications are rampant. In recent years, public opinion information about Android application security and illegal issues has continued to increase around the world, and domestic and foreign security agencies, media, government departments, etc. have all released reports on Android application threats.

[0003] However, the public opinion information on security threats released by different institutions or platforms often has different levels of structuring, and some sources lack authority and cannot provide effective support for corporate decision-making. In addition, due to the timeliness and complexity of global public opinion, companies cannot simply rely on manual public opinion screening and information analysis. Therefore, there is an urgent need for an automated system that can filter out highly relevant, authoritative and business-valuable information from massive public opinion information to provide support for companies and regulators. Faced with a large number of public opinion reports on application security at home and abroad, the existing system lacks an effective method to automatically filter out reports that are both thematically relevant and highly disseminated from these reports. Summary of the invention

[0004] The embodiments of the present invention provide a method, device, storage medium and terminal for recommending public opinion information. By acquiring a large amount of public opinion information related to the target topic, the value of each piece of public opinion information is scored from three aspects: relevance dimension, source dimension and dissemination dimension. According to the ranking of all public opinion information, public opinion information is recommended to enterprises.

[0005] In a first aspect, an embodiment of the present invention provides a method for recommending public opinion information, comprising:

[0006] Determine a target public opinion topic, and obtain a public opinion information set according to the target public opinion topic; the public opinion information set includes one or more public opinion information;

[0007] Obtaining value information of each piece of public opinion information in the public opinion information set; the value information includes one or more of relevance dimension information, source dimension information, and propagation dimension information;

[0008] Determine the score of any dimensional information according to the scoring rule corresponding to any dimensional information, thereby obtaining one or more dimensional information scores of each piece of public opinion information;

[0009] Based on one or more dimensional information scores of each piece of public opinion information and the weight coefficient of each dimensional information, the value score of each piece of public opinion information is calculated, and the public opinion information is recommended to the target public opinion topic according to the value score ranking of all public opinion information in the public opinion information set.

[0010] In a second aspect, an embodiment of the present invention provides a public opinion information recommendation device, comprising:

[0011] A public opinion information acquisition module, used to determine a target public opinion topic, and acquire a public opinion information set according to the target public opinion topic; the public opinion information set includes one or more public opinion information;

[0012] A value information acquisition module, used to acquire value information of each piece of public opinion information in the public opinion information set; the value information includes one or more of relevance dimension information, source dimension information and propagation dimension information;

[0013] A dimension information scoring module, used to determine the score of any dimension information according to the scoring rules corresponding to any dimension information, so as to obtain one or more dimension information scores of each piece of public opinion information;

[0014] The public opinion information recommendation module is used to calculate the value score of each piece of public opinion information based on one or more dimensional information scores of each piece of public opinion information and the weight coefficient of each dimensional information, and recommend public opinion information to the target public opinion topic according to the value score ranking of all public opinion information in the public opinion information set.

[0015] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method steps described in the first aspect are implemented.

[0016] In a fourth aspect, an embodiment of the present invention provides a terminal, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method steps described in the first aspect when executing the computer program.

[0017] The public opinion information recommendation method, device, storage medium and terminal provided by the embodiments of the present invention have the following technical effects:

[0018] After the embodiment of the present invention determines the target public opinion topic, one or more public opinion information related to the target public opinion topic is obtained to form a public opinion information set; one or more of the relevance dimension information, source dimension information and propagation dimension information of each piece of public opinion information in the public opinion information set is obtained; according to the scoring rules corresponding to each dimension information, the score of each dimension information is determined, and the value score of each piece of public opinion information is calculated in combination with the weight of each dimension information. Public opinion information is recommended based on the value score ranking of all public opinion information in the public opinion information set. By applying the above method, it is possible to automatically analyze and screen public opinion information from multiple dimensions, improve the efficiency and authority of public opinion information acquisition, and support real-time updates, so that enterprises or governments can obtain the latest public opinion information in real time, quickly respond, and avoid potential business risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 A flowchart of a method for recommending public opinion information provided by an embodiment of the present invention;

[0021] Figure 2 A schematic diagram of the structure of a public opinion information recommendation device provided by an embodiment of the present invention;

[0022] Figure 3 A block diagram of a terminal provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are only examples of devices and methods consistent with some aspects of the present invention as detailed in the attached claims.

[0025] In the description of the present invention, it should be understood that the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0026] In recent years, public opinion information about Android application security and illegal issues has continued to increase worldwide, and domestic and foreign security agencies, media, government departments, etc. have been issuing reports on Android application threats.

[0027] However, the public opinion information on security threats released by different institutions or platforms often has different levels of structuring, and some sources lack authority and cannot provide effective support for corporate decision-making. In addition, due to the timeliness and complexity of global public opinion, companies cannot simply rely on manual public opinion screening and information analysis. Therefore, there is an urgent need for an automated system that can filter out highly relevant, authoritative and business-valuable information from massive public opinion information to provide support for companies and regulators. Faced with a large number of public opinion reports on application security at home and abroad, the existing system lacks an effective method to automatically filter out reports that are both thematically relevant and highly disseminated from these reports.

[0028] In view of the above problems, the present invention provides a method, device, storage medium and terminal for recommending public opinion information. In the following, each step of the method for recommending public opinion information in the embodiment of the present invention will be described in more detail with reference to the accompanying drawings and embodiments.

[0029] Figure 1 A flowchart of a method for recommending public opinion information provided by an embodiment of the present invention. Figure 1 As shown, the method of the embodiment of the present invention may include the following steps:

[0030] S101, determine a target public opinion topic, and obtain a public opinion information set according to the target public opinion topic; the public opinion information set includes one or more public opinion information.

[0031] In the embodiment of the present invention, the target public opinion topic needs to be determined first, that is, the public opinion topic for which relevant public opinion information is to be obtained. The specific topic can be determined according to the needs. For example, with the development of the Internet, some emerging high-risk applications have emerged, such as gambling applications, illegal dating applications, etc. These applications not only violate laws and regulations, but are also prone to gambling fraud, privacy infringement, lack of protection for underage users, etc., causing great harm to users and society.

[0032] If you want to obtain relevant public opinion information about football gambling applications, you can identify them as the target public opinion topic. And according to the target public opinion topic, obtain multiple relevant public opinion information from public channels to form a public opinion information set. Public channels can be announcements and reports issued by government / official agencies, such as announcements and penalty notices issued by authoritative agencies such as the Cyberspace Administration of China and the Ministry of Industry and Information Technology; they can also be reports and related reports published by mainstream news media at home and abroad on news websites; you can also obtain relevant public opinion information through industry platforms, such as industry reports released by mobile phone manufacturers, security announcements and review information of corresponding application markets, etc.; you can also obtain relevant information from social media and self-media platforms, such as Weibo, forums, blogs and other self-media platforms to capture user comments and security feedback information; and some relatively niche special social platforms, such as Telegram groups, dark web forums, etc. Of course, the channels are not limited to the above descriptions, and relevant information can also be obtained through other channels.

[0033] After obtaining multiple pieces of public opinion information on the target public opinion topic, the public opinion information can be preprocessed, such as removing duplicates and clutter, to facilitate subsequent calculations. The preprocessed public opinion information is combined into a public opinion information set.

[0034] S102, obtaining value information of each piece of public opinion information in the public opinion information set; the value information includes one or more of relevance dimension information, source dimension information and propagation dimension information.

[0035] In an embodiment of the present invention, after determining the target public opinion topic and obtaining the public opinion information set related to the target public opinion topic, since the public opinion information set obtained through multiple channels includes a large amount of public opinion information, and the sources and types are various, there is a large amount of public opinion information with low relevance and accuracy, which cannot provide much help for understanding the current public opinion status of the target public opinion topic. Therefore, it is necessary to screen the public opinion information in the public opinion information set to obtain public opinion information with high relevance, high accuracy and practical value.

[0036] Specifically, the value information of each piece of public opinion information in the public opinion information set is obtained, and the value of each piece of public opinion information is judged by the value information to screen the public opinion information. The value information includes one or more of the relevance dimension information, the source dimension information and the propagation dimension information.

[0037] Optionally, the relevance dimension information is the relevant information content of each piece of public opinion information and the target public opinion theme, which can be specifically reflected in the words and sentences related to the target public opinion theme included in the public opinion information. For example, if the target public opinion theme is a football betting application, the public opinion information includes words such as football betting, football betting platform, gambling, and betting, which can be regarded as relevant words and serve as relevance dimension information. The more words and sentences related to the target public opinion theme included in the public opinion information, the stronger the relevance to the target public opinion theme, and the more help it can provide for analyzing the target public opinion theme. Therefore, the relevance dimension information of public opinion information plays an important role in the screening and recommendation of public opinion information.

[0038] Optionally, the source dimension information is the source of each piece of public opinion information, that is, the relevant information generated during the publishing process, including one or more of the source category, publisher qualifications, and review mechanism.

[0039] Optionally, public opinion information from different source categories has different credibility and influence in dissemination. For example, public opinion information released by the government / official agencies is more authoritative and influential than that released by mainstream news media or industry platforms, and public opinion information released by mainstream news media or industry platforms is more authoritative and influential than that released by relatively niche platforms such as self-media, personal blogs, complaint platforms, and special communities. Therefore, the source category is an important factor in judging the authority and influence of public opinion information.

[0040] Optionally, the source category includes one or more of the following: official agencies, mainstream news media, industry authoritative platforms, and others. Official agencies are government or some international professional agencies, such as China's Cyberspace Administration of China and the Ministry of Industry and Information Technology, or internationally, Google Security Bulletin and the EU Digital Market Regulator. Mainstream news media include domestic and foreign mainstream media (such as Xinhua News Agency, etc.), which have professional review and release mechanisms, and their information accuracy and credibility are high, but slightly inferior to government / official agencies. Industry authoritative platforms refer to some industry leading companies, such as Huawei, Xiaomi, OPPO, vivo app stores and industry report release platforms. The information released by such platforms has a high influence in the industry, especially reports on application security and compliance. The public opinion information released by relatively niche platforms such as self-media, personal blogs, complaint platforms, and special communities is less authoritative and can be classified as others. Of course, other smaller platforms can also be included.

[0041] Optionally, the publisher's qualifications determine the credibility of its information release. For example, the publisher's certification level can reflect the publisher's qualifications. Government agencies or industry certification platforms usually have higher credibility. At the same time, industry ranking is also an important factor in measuring the publisher's qualifications. In an industry, objects with higher industry rankings usually have higher authority and credibility. In addition, years of experience are also an important factor in measuring the publisher's qualifications. The longer the years of experience, the longer the publisher has been active in the field and the more information he knows. Long-term active publishers usually have higher qualifications. Therefore, publisher qualifications include one or more of the following: certification level, industry ranking rate, and years of experience.

[0042] Optionally, whether the published content has been through a strict review mechanism is also an important factor in measuring its authority. For example, government announcements usually go through multiple levels of review to ensure the accuracy and timeliness of the information. Self-media or personal blogs may lack an effective review mechanism. Therefore, the review mechanism is also an important factor in judging the authority and credibility of public opinion information. The review mechanism can be demonstrated by the number of independent verifications, which refers to whether the publisher provides independent verification or third-party verification support.

[0043] Optionally, the dissemination dimension information is the dissemination channel, dissemination method and dissemination scope of each piece of public opinion information. For example, a piece of public opinion information can be disseminated through news websites or social media. A large dissemination scope indicates a wide audience range, reflecting the scope of influence of the public opinion information. Specifically, it includes one or more of the acceptance degree and dissemination factors.

[0044] Optionally, the degree of acceptance refers to the audience's contact and interaction during the dissemination process of public opinion information. It also reflects the dissemination scope and popularity of public opinion information, which can be measured by the number of readings, comments and reposts of public opinion information.

[0045] Optional, the dissemination factor is a parameter for evaluating the effect of information dissemination, which may include dissemination channels and dissemination methods. Dissemination channels refer to the channels through which public opinion information is disseminated. Different channels have different dissemination effects. For example, social media, Weibo, WeChat, etc. have large user traffic, a large audience group, a relatively high dissemination range and speed, and are generally aimed at ordinary users; mainstream news websites, such as People's Daily Online and Huanqiu.com, have great influence, but the speed of information dissemination is slow, the scope is small, and the interactivity is relatively weak; industry forums, such as Zhihu, Hupu, CSDN, etc., are professional, but the audience is relatively concentrated and the dissemination range is limited; there are also some niche channels, such as blogs, private channels, etc., which have very weak dissemination range and speed, and can be uniformly classified as others.

[0046] The dissemination method determines the dissemination depth of information. Different dissemination forms have different dissemination strengths and influences. Dissemination forms can be divided into original and reprinted forms. Original content is generally more persuasive and disseminable than reprinted content, and can be further disseminated as it is quoted and forwarded. Reprinting means forwarding original content, adding some of your own opinions and views. According to the forwarding opinions, reprinting can also be divided into deep reprinting and short reprinting. If you express your own opinions in depth on the basis of forwarding an original content, it can be defined as deep reprinting. If you only express your opinions briefly or reprint unintentionally, it can be defined as short reprinting. Specifically, it can be divided according to the length of the text published in the reprint. If the length of the published text is greater than 100, 200, etc., it is classified as deep reprinting, otherwise it is short reprinting. The specific threshold can be determined according to the situation. Of course, it can also be divided according to other methods, such as according to the number of words related to the target public opinion theme contained in the reprinted text.

[0047] Based on any of the above optional embodiments, S103, determine the score of any dimensional information according to the scoring rule corresponding to any dimensional information, so as to obtain one or more dimensional information scores for each piece of public opinion information.

[0048] In the embodiment of the present invention, after obtaining one or more dimensional information of each piece of public opinion information in the public opinion information set, the situation of each dimensional information is different, so different scoring rules need to be used to score each dimensional information. Therefore, for any dimensional information, the score of the any dimensional information is determined according to the scoring rule corresponding to the any dimensional information.

[0049] In an optional embodiment, S103.1, with respect to the correlation dimension information, each piece of public opinion information is segmented to obtain a segmentation set; the segmentation set includes one or more segmentations.

[0050] Optionally, as can be seen from the foregoing, the relevance dimension information is the relevant information content of each piece of public opinion information and the target public opinion topic, which can be specifically reflected in the words and sentences related to the target public opinion topic included in the public opinion information. Therefore, when calculating the similarity between each piece of public opinion information and the target public opinion topic, each piece of public opinion information is first segmented, and the obtained one or more segmented words are combined into a segmented word set.

[0051] Specifically, public opinion information generally includes a lot of irrelevant information, so first clean the public opinion information and remove irrelevant noise words, such as common stop words and advertising words, to reduce interference with subsequent analysis. Then normalize the text, including case conversion, removal of punctuation and special characters, etc., to ensure the consistency of the text. Perform word segmentation on the public opinion information to split the text into vocabulary units. Tools that can be used include NLTK, spaCy, Jieba, etc. And restore the vocabulary to its root form to reduce the impact of vocabulary changes on the analysis. Root extraction tools can use Porter Stemmer or Lancaster Stemmer, etc.

[0052] S103.2, based on the word set, the subject word library and the public opinion information set, calculate the relevance score between each piece of public opinion information and each keyword in the subject word library; the subject word library includes one or more keywords related to the target public opinion topic.

[0053] Optionally, after obtaining the word segmentation set of each public opinion information, the relevance score of each public opinion information and each keyword in the theme word library is calculated based on the word segmentation set, the theme word library and the public opinion information set; wherein the theme word library includes one or more keywords related to the target public opinion topic. The theme word library can be determined according to the target public opinion topic, and must contain core vocabulary, synonyms and related terms for the target public opinion topic to ensure that it can cover a wide range of subject content and form a theme word family. For example, the core vocabulary of the gambling subject content is: gambling, illegal World Cup betting, and the synonyms and related terms are: gambling, gambling platform, betting, gambling website, gambling software, gambling betting, gambling cheating, gambling fraud, gambling activities, black market gambling, illegal gambling behavior, illegal gambling.

[0054] Specifically, the relevance score between each piece of public opinion information and each keyword in the subject word library is obtained according to the following formula:

[0055]

[0056] Among them, d is each of the above public opinion information, t is each of the above keywords, B(d,t) is the correlation score between each of the above public opinion information and each of the above keywords, TF(d,t) is the number of times each of the above keywords appears in each of the above public opinion information, |d| is the number of segmentations in the segmentation set of each of the above public opinion information, and avgdl is the average |d| of all public opinion information in the public opinion information set.

[0057] The IDF(t) is obtained according to the following formula:

[0058]

[0059] Among them, N is the number of public opinion information in the public opinion information set, and DF(t) is the number of public opinion information including keyword t in the public opinion information set.

[0060] S103.3, taking the sum of the relevance scores between each piece of public opinion information and all keywords in the subject word library as the relevance dimension score between each piece of public opinion information and the target public opinion topic.

[0061] Optionally, after calculating the relevance score between each piece of public opinion information and each keyword in the subject vocabulary according to the above method, since the subject vocabulary includes one or more keywords, one or more relevance scores between each piece of public opinion information and the subject vocabulary can be obtained, and the sum of all relevance scores can be used as the relevance dimension score between each piece of public opinion information and the target public opinion topic.

[0062] In another optional embodiment, S103.1, for the source dimension information, one or more information scores are obtained according to a scoring rule for each source dimension information;

[0063] S103.2, based on the one or more information scores and the weight coefficient of each information, calculate the source dimension score of each piece of public opinion information.

[0064] Optionally, as can be seen from the above, the source dimension information includes one or more of source category, publisher qualifications and review mechanism; among which, the source category includes one or more of official agencies, mainstream news media, industry authoritative platforms and others; the publisher qualifications include one or more of certification level, industry ranking rate, and years of experience; the review mechanism includes the number of independent verifications.

[0065] As can be seen from the above, the source categories, publisher qualifications and review mechanisms, including the content and format of the information, are different, and their scoring rules are also different. Therefore, when scoring the source dimension information, it is necessary to score according to the scoring rules of each source dimension information, and one or more information scores can be obtained. Then, based on one or more information scores and the weight coefficient of each information, the source dimension score of each public opinion information is calculated.

[0066] Optionally, the scoring rules corresponding to the source category include:

[0067] If the source category is an official agency, the score is A1;

[0068] If the source category is mainstream news media, the score is A2;

[0069] If the source category is an industry authority platform, the score is A3;

[0070] If the source category is other, the score is A4;

[0071] Among them, A1>A2>A3>A4.

[0072] As can be seen from the above, the source categories include one or more of official agencies, mainstream news media, industry authoritative platforms and others. Official agencies are government or some international professional agencies, such as China's Cyberspace Administration of China and the Ministry of Industry and Information Technology, or internationally, Google Security Notices and EU Digital Markets Regulatory Departments, etc. They have the greatest authority and influence and should be scored the highest, such as 4 points; mainstream news media include domestic and foreign mainstream media (such as Xinhua News Agency, etc.). These media have professional review and release mechanisms, and their information accuracy and credibility are relatively high, but they are slightly inferior to government / official agencies. Therefore, the score should be slightly lower than that of official agencies, and can be rated 3 points. Industry authoritative platforms refer to some industry leading companies, such as Huawei, Xiaomi, OPPO, vivo app stores and industry report release platforms. The information released by such platforms has a high influence in the industry, but the authority is lower than that of mainstream news media, so the score should be slightly lower than that of mainstream news media, and can be rated 2 points. The public opinion information released by relatively niche platforms such as self-media, personal blogs, complaint platforms and special communities is less authoritative and can be classified as others. This type of source has the lowest authority, accuracy and credibility, so it has the lowest score and can be rated 1 point. Of course, the specific scoring mechanism can be set according to actual conditions and is not limited here.

[0073] Optionally, the publisher's qualifications include one or more of the following: certification level, industry ranking rate, and years of experience. Certification levels are generally divided into four levels: Level 1 certification, certified by national or international authoritative certification agencies, such as the Cyberspace Administration of China, the Ministry of Industry and Information Technology, and the International Organization for Standardization ISO; Level 2 certification: certification by industry associations or authoritative organizations, such as the China Information Security Evaluation Center and well-known domestic industry associations; Level 3 certification: local or internal industry certification, such as certification by local government departments or industry associations in specific fields; Uncertified: No industry certification, but may have certain influence or qualifications. Therefore, the scoring rules can be:

[0074] If the certification level is Level 1, the score is B1;

[0075] If the certification level is Level 2, the score is B2;

[0076] If the certification level is level 3, the score is B3;

[0077] If the certification level is uncertified, the score is B4;

[0078] Among them, B1>B2>B3>B4. B1, B2, B3, and B4 can be determined according to actual conditions and are not limited here.

[0079] At the same time, industry ranking is also an important factor in measuring the qualifications of publishers. In an industry, objects with higher industry rankings have higher authority and credibility. However, industry rankings are from the first to the last. If industry rankings are used as the calculation standard, publishers with lower rankings will have higher scores. Therefore, the industry ranking rate, that is, the ratio of the publisher's ranking in the industry to the total number of rankings, is used as the score. The years of employment are specific numbers and can be used directly as the score. Then, one or more of the certification level score, industry ranking rate, and years of employment are weighted to obtain the score of the publisher's qualifications. The specific weight coefficient can be determined according to the actual situation and is not limited here.

[0080] Optionally, the audit mechanism is displayed through the number of independent verifications. Independent verification refers to whether the publisher provides independent verification or third-party verification support. The number of independent verifications is a specific number and can be directly used as a score.

[0081] According to the above method, after obtaining the score of each information in each source dimension information, the source dimension score of each public opinion information can be calculated according to one or more information scores and the weight coefficient of each information. The specific weight coefficient can be determined according to the actual situation and is not limited here.

[0082] In another optional embodiment, S103.1, for the dissemination dimension information, according to the scoring rule of each dissemination dimension information, obtain the acceptance degree score and the dissemination factor score;

[0083] S103.2, taking the product of the acceptance level score and the communication factor score as the communication dimension score of each piece of public opinion information.

[0084] Optionally, as can be seen from the above, the communication dimension information includes one or more of the degree of acceptance and communication factors; wherein the degree of acceptance includes one or more of the number of readings, the number of comments and the number of reposts; and the communication factors include one or more of the communication channels and the communication methods.

[0085] As can be seen from the above, the acceptance degree and communication factors include different information content and formats, and their scoring rules are also different. Therefore, when scoring the communication dimension information, it is necessary to score according to the scoring rules of each source dimension information to obtain one or more of the acceptance degree score and the communication factor score. The product of the acceptance degree score and one or more of the communication factor scores is used as the communication dimension score of each public opinion information. If there is only an acceptance degree score or a communication factor score, the acceptance degree score or the communication factor score is used as the communication dimension score of each public opinion information.

[0086] Optionally, the scoring rules corresponding to the acceptance level include:

[0087] A weighted calculation is performed on one or more of the reading volume, the number of comments, and the number of reposts to obtain a score of the acceptance level.

[0088] The number of reads, comments and reposts are all specific numbers and can be calculated directly. The specific weight coefficient can be determined according to the actual situation and is not limited here.

[0089] Optionally, the scoring rules corresponding to the propagation factors include:

[0090] If the communication channel is social media, the score is C1;

[0091] If the dissemination channel is a mainstream news website, the score is C2;

[0092] If the communication channel is an industry channel, the score is C3;

[0093] If the communication channel is other, the score is C4;

[0094] Among them, C1>C2>C3>C4.

[0095] The dissemination channel refers to the channel through which public opinion information is disseminated. Different channels have different dissemination effects. For example, social media, Weibo, etc. have large user traffic, a large audience group, a relatively high dissemination range and speed, and the highest score; mainstream news websites, such as People's Daily Online and Huanqiu.com, have great influence, but the information dissemination speed is slow, the scope is small, and the interactivity is relatively weak, and the score is slightly lower than social media; industry forums, such as Zhihu, Hupu, CSDN, etc., are professional, but the target groups are relatively concentrated, the dissemination range is limited, and the score is second; there are also some niche channels, such as blogs, private channels, etc., which have very weak dissemination range and speed, and can be uniformly classified as others, with the lowest score. C1, C2, C3, and C4 can be determined according to actual conditions and are not limited here.

[0096] Optional, for the transmission method:

[0097] If the dissemination method described is original content, the score is D1;

[0098] If the dissemination method is in-depth reprinting, the score is D2;

[0099] If the dissemination method is brief reprint, the score is D3;

[0100] Among them, D1>D2>D3.

[0101] The dissemination method determines the depth of information dissemination. Different dissemination forms have different dissemination strengths and influences. Dissemination forms can be original or reprinted. Original content has independent information sources, high credibility and influence, can attract more attention, and has the highest score. Deep reprinting is based on the deep processing of original content, has strong ductility and depth, is easy to trigger discussion, and has the second highest score than original content. Brief reprinting is a brief processing of original content, which is not very discussed and has the lowest score. D1, D2, and D3 can be determined according to actual conditions and are not limited here.

[0102] According to the above method, after obtaining the communication channel score and the communication mode score, one or more can be selected for weighted calculation to obtain the communication factor score. The specific weight coefficient can be determined according to the actual situation and is not limited here.

[0103] Based on any of the above optional embodiments, S104, based on one or more dimensional information scores of each piece of public opinion information and the weight coefficient of each dimensional information, calculate the value score of each piece of public opinion information, and recommend public opinion information according to the value score ranking of all public opinion information in the public opinion information set.

[0104] In an embodiment of the present invention, after obtaining one or more dimensional information scores based on one or more dimensional information among the relevance dimensional information, source dimensional information and propagation dimensional information of each piece of public opinion information, a weighted calculation is performed in combination with the weight coefficient of each dimensional information to obtain the value score of each piece of public opinion information.

[0105] Similarly, the value scores of all public opinion information in the public opinion information set can be obtained, and then the public opinion information can be recommended according to the value score ranking of all public opinion information in the public opinion information set. The value scores of all public opinion information in the public opinion information set can be ranked from high to low, and the top 3 or top 5 public opinion information in the value score ranking can be selected for recommendation. The specific selection of the top few public opinion information in the value score ranking for recommendation can be determined according to actual conditions and is not limited here.

[0106] It should be noted that the above embodiment only targets one target public opinion topic, but in actual situations, it may be necessary to recommend public opinion information for multiple public opinion topics. Therefore, multiple public opinion topics are selected, and the public opinion information set corresponding to each public opinion topic is obtained, and the value score of each public opinion in the public opinion information set is calculated to form a corresponding value score set. The public opinion information sets and value score sets corresponding to multiple public opinion topics form a public opinion information recommendation system. In order to recommend public opinion information at any time.

[0107] In summary, an embodiment of the present invention provides a method for recommending public opinion information. After determining a target public opinion topic, one or more public opinion information related to the target public opinion topic is obtained to form a public opinion information set; one or more of the relevance dimension information, source dimension information, and propagation dimension information of each piece of public opinion information in the public opinion information set is obtained; according to the scoring rules corresponding to each dimension information, the score of each dimension information is determined, and the value score of each piece of public opinion information is calculated in combination with the weight of each dimension information. Public opinion information is recommended according to the value score ranking of all public opinion information in the public opinion information set. By applying the above method, it is possible to automatically analyze and screen public opinion information from multiple dimensions, improve the efficiency and authority of obtaining public opinion information, and support real-time updates, so that enterprises or governments can obtain the latest public opinion information in real time, quickly respond, and avoid potential business risks.

[0108] The following are embodiments of the apparatus of the present invention, which can be used to implement the embodiments of the method of the present invention. For details not disclosed in the embodiments of the apparatus of the present invention, please refer to the embodiments of the method of the present invention.

[0109] See also Figure 2 , which is a structural diagram of a public opinion information recommendation device provided in an embodiment of the present invention.

[0110] The public opinion information recommendation device 200 in the embodiment of the present invention includes: a public opinion information acquisition module 201, a value information acquisition module 202, a dimension information scoring module 203 and a public opinion information recommendation module 204.

[0111] A public opinion information acquisition module, used to determine a target public opinion topic, and acquire a public opinion information set according to the target public opinion topic; the public opinion information set includes one or more public opinion information;

[0112] A value information acquisition module, used to acquire value information of each piece of public opinion information in the public opinion information set; the value information includes one or more of relevance dimension information, source dimension information and propagation dimension information;

[0113] A dimension information scoring module, used to determine the score of any dimension information according to the scoring rules corresponding to any dimension information, so as to obtain one or more dimension information scores of each piece of public opinion information;

[0114] The public opinion information recommendation module is used to calculate the value score of each piece of public opinion information based on one or more dimensional information scores of each piece of public opinion information and the weight coefficient of each dimensional information, and recommend public opinion information to the target public opinion topic according to the value score ranking of all public opinion information in the public opinion information set.

[0115] It should be noted that the public opinion information recommendation device provided in the above embodiment only uses the division of the above functional modules as an example when executing the public opinion information recommendation method. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the public opinion information recommendation device provided in the above embodiment belongs to the same concept as the public opinion information recommendation method embodiment. Therefore, for details not disclosed in the embodiment of the device of the present invention, please refer to the embodiment of the public opinion information recommendation method of the present invention, which will not be repeated here.

[0116] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0117] The embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, the steps of the method of any of the above embodiments are implemented. The computer-readable storage medium may include, but is not limited to, any type of disk, including a floppy disk, an optical disk, a DVD, a CD-ROM, a micro drive, and a magneto-optical disk, a ROM, a RAM, an EPROM, an EEPROM, a DRAM, a VRAM, a flash memory device, a magnetic card or an optical card, a nanosystem (including a molecular memory IC), or any type of medium or device suitable for storing instructions and / or data.

[0118] An embodiment of the present invention further provides a terminal, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of any of the above-mentioned method embodiments when executing the program.

[0119] Figure 3 A block diagram of a terminal provided by an embodiment of the present invention. Figure 3, the embodiment of the present invention provides a terminal 300, including: a processor (processor) 301, a communication interface (Communications Interface) 302, a memory (memory) 303 and a communication bus 304, wherein the processor 301, the communication interface 302, and the memory 303 complete mutual communication through the communication bus 304. The processor 301 can call the logic instructions in the memory 303 to execute the following method, including: determining a target public opinion topic, and obtaining a public opinion information set according to the target public opinion topic; the public opinion information set includes one or more public opinion information; obtaining value information of each public opinion information in the public opinion information set; the value information includes one or more of correlation dimension information, source dimension information and propagation dimension information; according to the scoring rule corresponding to any dimension information, determining the score of any dimension information, thereby obtaining one or more dimension information scores of each public opinion information; based on the one or more dimension information scores of each public opinion information and the weight coefficient of each dimension information, calculating the value score of each public opinion information, and recommending public opinion information according to the value score ranking of all public opinion information in the public opinion information set.

[0120] The terminal structure block diagram shown in the embodiment of the present invention does not constitute a limitation on the terminal 300. The terminal 300 may include more or fewer components than those shown in the figure, or combine certain components, or adopt a different component arrangement.

[0121] An embodiment of the present invention discloses a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the methods provided by the above-mentioned method embodiments, for example, including: determining a target public opinion topic, and obtaining a public opinion information set based on the target public opinion topic; the public opinion information set includes one or more public opinion information; obtaining value information of each public opinion information in the public opinion information set; the value information includes one or more of relevance dimension information, source dimension information and propagation dimension information; determining a score of any dimensional information according to a scoring rule corresponding to any dimensional information, thereby obtaining one or more dimensional information scores for each public opinion information; calculating the value score of each public opinion information based on the one or more dimensional information scores of each public opinion information and the weight coefficient of each dimensional information, and recommending public opinion information according to the value score ranking of all public opinion information in the public opinion information set.

[0122] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for recommending public opinion information, characterized in that: include: Determine a target public opinion topic, and obtain a public opinion information set according to the target public opinion topic; The public opinion information set includes one or more pieces of public opinion information; Obtaining value information of each piece of public opinion information in the public opinion information set; the value information includes one or more of relevance dimension information, source dimension information, and propagation dimension information; Determine the score of any dimensional information according to the scoring rule corresponding to any dimensional information, thereby obtaining one or more dimensional information scores of each piece of public opinion information; Based on the one or more dimensional information scores of each piece of public opinion information and the weight coefficient of each dimensional information, the value score of each piece of public opinion information is calculated, and public opinion information is recommended according to the value score ranking of all public opinion information in the public opinion information set.

2. The method according to claim 1, characterized in that: Determining the score of any dimensional information according to the scoring rule corresponding to any dimensional information, thereby obtaining one or more dimensional information scores of each piece of public opinion information, includes: According to the relevance dimension information, each piece of public opinion information is segmented to obtain a segmentation set; the segmentation set includes one or more segmentations; Based on the word set, the subject word library and the public opinion information set, calculating the relevance score between each piece of public opinion information and each keyword in the subject word library; the subject word library includes one or more keywords related to the target public opinion topic; The sum of the relevance scores between each piece of public opinion information and all keywords in the subject word library is used as the relevance dimension score between each piece of public opinion information and the target public opinion topic.

3. The method according to claim 2, characterized in that: The relevance score of each piece of public opinion information and each keyword in the subject word library is obtained according to the following formula: Wherein, d is each piece of public opinion information, t is each keyword, B(d, t) is the correlation score between each piece of public opinion information and each keyword, TF(d, t) is the number of times each keyword appears in each piece of public opinion information, |d| is the number of word segments in the word segmentation set of each piece of public opinion information, and avgdl is the average of |d| of all public opinion information in the public opinion information set; The IDF(t) is obtained according to the following formula: Among them, N is the number of public opinion information in the public opinion information set, and DF(t) is the number of public opinion information including keyword t in the public opinion information set.

4. The method according to claim 1, characterized in that: The source dimension information includes one or more of source category, publisher qualification and review mechanism; The source categories include one or more of the following: official agencies, mainstream news media, industry authoritative platforms and others; The publisher qualifications include one or more of the following: certification level, industry ranking rate, and years of experience; Said audit mechanism includes independent verification of quantity; Determining the score of any dimensional information according to the scoring rule corresponding to any dimensional information, thereby obtaining one or more dimensional information scores of each piece of public opinion information, includes: For the source dimension information, one or more information scores are obtained according to the scoring rules for each source dimension information; Based on the one or more information scores and the weight coefficient of each information, a source dimension score of each piece of public opinion information is calculated.

5. The method according to claim 4, characterized in that: The scoring rules corresponding to the source categories include: If the source category is an official agency, the score is A1; If the source category is mainstream news media, the score is A2; If the source category is an industry authority platform, the score is A3; If the source category is Other, the score is A4; Among them, A1>A2>A3>A4; The scoring rules corresponding to the publisher qualifications include: If the certification level is Level 1, the score is B1; If the certification level is Level 2, the score is B2; If the certification level is level 3, the score is B3; If the certification level is uncertified, the score is B4; Among them, B1>B2>B3>B4; The industry ranking rate is the ratio of the publisher’s ranking in the industry to the total number of rankings; Performing weighted calculation on one or more of the certification level score, the industry ranking rate and the years of experience to obtain a score for the publisher's qualifications; The scoring rules corresponding to the review mechanism include: The number of independent verifications is obtained as a score for the audit mechanism.

6. The method according to claim 1, characterized in that: The communication dimension information includes one or more of acceptance degree and communication factors; The acceptance level includes one or more of the following: number of readings, number of comments, and number of reposts; The communication factors include one or more of the following: communication channels and communication methods; Determining the score of any dimensional information according to the scoring rule corresponding to any dimensional information, thereby obtaining one or more dimensional information scores of each piece of public opinion information, includes: For the communication dimension information, according to the scoring rules of each communication dimension information, one or more of the acceptance degree score and the communication factor score are obtained; The product of one or more of the acceptance score and the propagation factor score is used as the propagation dimension score of each piece of public opinion information.

7. The method according to claim 6, characterized in that: The scoring rules corresponding to the acceptance level include: Perform weighted calculation on one or more of the reading volume, the number of comments, and the number of reposts to obtain the acceptance score: The scoring rules corresponding to the propagation factors include: If the communication channel is social media, the score is C1; If the dissemination channel is a mainstream news website, the score is C2; If the communication channel is an industry channel, the score is C3; If the communication channel is other, the score is C4; Among them, C1>C2>C3>C4; If the dissemination method described is original content, the score is D1; If the dissemination method is in-depth reprinting, the score is D2; If the dissemination method is brief reprint, the score is D3; Among them, D1>D2>D3; A weighted calculation is performed on one or more of the communication channel score and the communication mode score to obtain a score of the communication factor.

8. A public opinion information recommendation device, characterized in that: include: A public opinion information acquisition module, used to determine a target public opinion topic and acquire a public opinion information set according to the target public opinion topic; The public opinion information set includes one or more pieces of public opinion information; A value information acquisition module, used to acquire value information of each piece of public opinion information in the public opinion information set; the value information includes one or more of relevance dimension information, source dimension information and propagation dimension information; A dimension information scoring module, used to determine the score of any dimension information according to the scoring rules corresponding to any dimension information, so as to obtain one or more dimension information scores of each piece of public opinion information; The public opinion information recommendation module is used to calculate the value score of each piece of public opinion information based on one or more dimensional information scores of each piece of public opinion information and the weight coefficient of each dimensional information, and recommend public opinion information to the target public opinion topic according to the value score ranking of all public opinion information in the public opinion information set.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

10. A terminal comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

Images