A method, system, electronic device and storage medium for determining hot topics

Abstract

The present invention relates to a method, system, electronic device and storage medium for determining hot topics. The determination method comprises: obtaining interactive text data from user interactive text data, grouping and aggregating the interactive text data according to entity names to obtain multiple groups of text data to be aggregated, and adding topic weight values ​​to topic keywords respectively according to the topic relationship between topic keywords in the text data to be aggregated, obtaining the total weight value of the topic keywords according to the topic weight values ​​of the topic keywords, and determining the hot topics in the topic keywords according to the total weight value of the topic keywords, so as to realize the aggregation of topic data hotly discussed by users in the station, so as to quickly determine the hot spots concerned by users in the station, and provide data support for the operation work of operators.

Description

Technical Field

[0001] The present invention relates to the field of data analysis technology, and in particular to a hot topic determination method, system, electronic device and storage medium. Background Art

[0002] Video websites produce tens of thousands of pieces of content data every day. By quickly aggregating the website's popular data and locating current hot spots, they can quickly and effectively summarize information on the same topic and form hot topics.

[0003] The text information on current video websites mainly includes comments, barrage, video titles and corresponding attribute information. For hot topics, many users will further confirm the corresponding topics through search after hearing the corresponding news. The existing hot topics are generally displayed directly in the search window. For websites, if the displayed topics do not match the data that users want to search, it will affect users' doubts about the authority of the website, and it will also fail to provide users with a good user experience. Therefore, for websites, how to analyze the current hot topics is an urgent problem to be solved. Summary of the invention

[0004] In order to solve the problem in the prior art of how to analyze and obtain current hot topics from interactive text data, at least one embodiment of the present invention provides a hot topic determination method, system, electronic device and storage medium.

[0005] In a first aspect, an embodiment of the present invention provides a method for determining a hot topic, and the method for obtaining the hot topic includes:

[0006] Acquire text data to be aggregated from the interactive text data; the heat value of the text data to be aggregated is higher than a preset threshold;

[0007] For each group of text data to be aggregated, extract topic keywords in the text data to be aggregated, and determine a topic weight value of each topic keyword according to the topic relationship between the topic keywords;

[0008] The topic weight values ​​of the topic keywords are aggregated to obtain a total weight value of each topic keyword, and the hot topics in the topic keywords are determined based on the total weight value of the topic keywords.

[0009] Based on the above technical solution, the embodiment of the present invention can also make the following improvements.

[0010] In combination with the first aspect, in a first embodiment of the first aspect, obtaining the text data to be aggregated from the interactive text data includes:

[0011] Obtaining popularity data of the interactive text data; the popularity data includes: at least one of the number of comments on the text data, the field to which the text data belongs, and the existence time of the text data;

[0012] According to a preset scoring standard, adding a corresponding score to the popularity data of each group of the interactive text data;

[0013] For each group of the interactive text data, calculating the heat value according to the score of the heat data of the interactive text data;

[0014] The interactive text data with a heat value higher than a preset threshold is used as the text data to be aggregated.

[0015] In combination with the first aspect, in a second embodiment of the first aspect, extracting topic keywords from the text data to be aggregated, and determining a topic weight value of each of the topic keywords according to the topic relationship between the topic keywords, includes:

[0016] Acquire sentence components in the text data to be aggregated;

[0017] Determining topic keywords in the text data to be aggregated according to the sentence components;

[0018] The grammatical relationship between the topic keywords is obtained, a first topic weight value is configured for the topic keywords with a grammatical relationship, and a second topic weight value is configured for the topic keywords without a grammatical relationship; the first topic weight value is greater than the second topic weight value.

[0019] In combination with the first aspect, in a third embodiment of the first aspect, extracting topic keywords from the text data to be aggregated, and determining a topic weight value of each of the topic keywords according to the topic relationship between the topic keywords, includes:

[0020] Comparing the source data of the text data to be aggregated with the preset special symbol to determine whether the source data contains the preset special symbol;

[0021] If the source data has a preset special symbol, the words corresponding to the preset special symbol in the text data to be aggregated are used as candidate keywords;

[0022] According to the number of repetitions of the candidate keyword in the text data to be aggregated, determining whether the text data to be aggregated is low-quality data;

[0023] When the text data to be aggregated corresponding to the candidate keyword is not low-quality data, the candidate keyword is a topic keyword, and a third topic weight value is added to the topic keyword;

[0024] When the text data to be aggregated corresponding to the candidate keyword is low-quality data, the candidate keyword is deleted.

[0025] In combination with the first aspect, in a fourth embodiment of the first aspect, extracting topic keywords from the text data to be aggregated, and determining a topic weight value of each of the topic keywords according to the topic relationship between the topic keywords, includes:

[0026] Inputting the text data to be aggregated into a pre-established relation extraction model to obtain all subjects in the text data to be aggregated and objects corresponding to each subject;

[0027] Inputting the subject and the pre-established grammatical relationship into a pre-established object prediction model to obtain a predicted object;

[0028] Determining whether the predicted object is similar to an object corresponding to the subject;

[0029] If the predicted object is similar to the object corresponding to the subject, the subject and the object are used as topic keywords, and a fourth topic weight value is added to the topic keyword;

[0030] If the predicted object is not similar to the object corresponding to the subject, then the subject has no grammatical relationship with the object.

[0031] In combination with the fourth embodiment of the first aspect, in a fifth embodiment of the first aspect, inputting the text data to be aggregated into a pre-established relationship extraction model to obtain all subjects in the text data to be aggregated and objects corresponding to each subject, includes:

[0032] According to a preset correspondence between characters and numbers, each character in the text data to be aggregated is encoded to obtain a digital sequence that is the same as the order of the characters in the text data to be aggregated;

[0033] The digital sequence is input into the relation extraction model to obtain all subjects in the text data to be aggregated and the objects corresponding to each subject.

[0034] In combination with the first aspect or the first, second, third or fourth embodiment of the first aspect, in a fifth embodiment of the first aspect, the determination method further includes:

[0035] Determine the information interface associated with each hot topic;

[0036] The hot topics are sorted according to the total weight values ​​corresponding to the hot topics, and the information interfaces corresponding to the hot topics are displayed in sequence.

[0037] In a second aspect, an embodiment of the present invention provides a hot topic determination system, the determination system comprising:

[0038] A first acquisition unit is used to acquire text data to be aggregated from the interactive text data; the heat value of the text data to be aggregated is higher than a preset threshold;

[0039] A second acquisition unit is used to extract topic keywords from each group of text data to be aggregated;

[0040] A weight determination unit, used to determine a topic weight value of each of the topic keywords according to the topic relationship between the topic keywords;

[0041] The determination unit is used to summarize the topic weight values ​​of each topic keyword to obtain the total weight value of each topic keyword; and determine the hot topics in the topic keywords according to the total weight value of the topic keywords.

[0042] In a third aspect, an embodiment of the present invention provides an electronic device, including a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus;

[0043] Memory, used to store computer programs;

[0044] The processor is used to implement the hot topic determination method described in any embodiment of the first aspect when executing the program stored in the memory.

[0045] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the hot topic determination method described in any embodiment of the first aspect.

[0046] The above-mentioned technical scheme of the present invention has the following advantages compared with the prior art: this scheme obtains interactive text data from user interaction text data, groups and aggregates the interactive text data to obtain text data to be aggregated, and adds topic weight values ​​to topic keywords respectively according to the topic relationship between topic keywords in the text data to be aggregated, obtains the total weight value of topic keywords according to the topic weight values ​​of topic keywords, and determines the hot topics in topic keywords according to the total weight values ​​of topic keywords, thereby realizing the aggregation of topic data hotly discussed by users on the site, so as to quickly determine the hot spots that users on the site are concerned about, and provide data support for the operation work of operators. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a flowchart of a method for determining a hot topic provided by an embodiment of the present invention;

[0048] Figure 2 is a flow chart of a method for determining a hot topic provided by another embodiment of the present invention;

[0049] Figure 3 This is a flowchart of a method for determining a hot topic provided by another embodiment of the present invention;

[0050] Figure 4 This is a second flow chart of a method for determining a hot topic provided by another embodiment of the present invention;

[0051] Figure 5 This is a third flow chart of a method for determining a hot topic provided by another embodiment of the present invention;

[0052] Figure 6 FIG4 is a flow chart of a method for determining a hot topic provided by another embodiment of the present invention;

[0053] Figure 7 is a schematic diagram of the structure of a hot topic determination system provided by another embodiment of the present invention;

[0054] Figure 8 This is a schematic diagram of the structure of an electronic device provided by another embodiment of the present invention. DETAILED DESCRIPTION

[0055] To make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0056] An embodiment of the present invention provides a structural diagram of a system for implementing various embodiments of the present invention, including a terminal, wherein the terminal includes: a processor 1110 and a memory 1130, wherein the terminal can be a mobile terminal such as a mobile phone, a tablet computer, a laptop computer, a handheld computer, a personal digital assistant (Personal Digital Assistant, PDA), a portable media player (Portable Media Player, PMP), a navigation device, a wearable device, a smart bracelet, a pedometer, etc., as well as a fixed terminal such as a digital TV, a desktop computer, etc.

[0057] like Figure 1 As shown, an embodiment of the present invention provides a method for determining a hot topic. Figure 1 As shown, the determination method includes the following steps:

[0058] S11. Obtain interactive text data from the user interaction text data; the heat value of the interactive text data is higher than a preset threshold.

[0059] In this embodiment, the video website produces tens of thousands of content data every day. By quickly aggregating the popular data of the website and locating the current hot spots, the same topic information can be quickly and effectively summarized to form hot topics. The text information of current video websites mainly includes comments, bullet screens, video titles and corresponding attribute information, among which the massive comments and bullet screens have strong user randomness and scattered key points.

[0060] In this embodiment, the user interaction text data is data generated by the interaction between users, such as bullet screens, comments, replies and other information, and the interactive text data is obtained from the user interaction text data. For example, the user interaction amount of each user interaction text data can be used as a heat value. Specifically, for example, the bullet screen on a video website is a kind of interactive text. In this scheme, the number of likes of the bullet screen can be used as a heat value, thereby determining the text data to be aggregated in the interactive text data. The heat coefficient can also be added according to the influence of the object involved in the user interaction text data, the user's comments, replies, content length and existence time, and the heat coefficient can be calculated. Statistics are performed to obtain the heat value of the user interaction text data. For example, the average content length can be determined based on historical data, and the content length of the user interaction text data can be divided by the average content length to obtain a corresponding heat coefficient. Then, the average existence time can be determined based on historical data, and the content length of the user interaction text data can be divided by the average existence time to obtain a corresponding heat coefficient. Several other parameters can also be used in the same way to achieve heat evaluation of data of different dimensions. By normalizing different parameters separately, the heat value of the user interaction text data can be determined, and the user interaction text data with a heat value higher than a preset threshold is used as interactive text data.

[0061] In this embodiment, since this scheme ultimately needs to determine hot topics, and the amount of user interaction text data is very large whether on video websites or other types of websites, in this step, we filter the interactive text data according to the heat value of the user interaction text data to obtain the interactive text data in this step, which can effectively reduce the amount of data processing and, at the same time, improve the effectiveness of the hot topics finally determined.

[0062] S12. Group and aggregate the interactive text data according to entity names to obtain multiple groups of text data to be aggregated.

[0063] In this embodiment, entities are names of people, names of organizations, names of places, and all other entities identified by names. More extensive entities also include numbers, dates, currencies, addresses, and the like. In this solution, since the interactive text data is content input by people in chat, the client or the system cannot directly understand the semantics expressed by the content. In this solution, the interactive text data is grouped according to the entity names in each interactive text data to obtain multiple groups of text data to be aggregated, thereby reducing the difficulty of analysis within each group of text data to be aggregated and improving processing efficiency.

[0064] For example, the interactive text data may be grouped according to the degree of repetition between entities contained in each interactive text data. Specifically, the interactive text data may be screened according to the number of repeated entity names between the interactive text data and a pre-set preset number. When the number of consistent entity names between two interactive text data is greater than the preset number, the two interactive text data are grouped together. The interactive text data may also be grouped according to the entity name. For example, the interactive text data containing the same entity name are regarded as the same group of text data to be aggregated. Of course, other methods may also be used to group and aggregate the interactive text data by entity name, and this scheme does not specifically limit this.

[0065] S13. For each group of text data to be aggregated, extract topic keywords in the text data to be aggregated, and add topic weight values ​​to the topic keywords respectively according to the topic relationships between the topic keywords.

[0066] In this embodiment, topic keywords in the text data to be aggregated are obtained, which may be sentence components such as the subject, predicate, object, etc. in the text data to be aggregated, or the text data to be aggregated may be processed through an existing semantic recognition model to extract topic keywords in the text data to be aggregated, for example, useful information in the data may be extracted through a text summary model.

[0067] In this embodiment, the topic relationship between topic keywords can also be obtained through relationship extraction or syntactic analysis. For example, relationship extraction can obtain the contextual relationship between different entities in a sentence, and syntactic analysis can be used to obtain the syntactic relationship between topic keywords, such as the subject-object relationship with postposition of the attributive, the subject-predicate-verb-complement relationship with preposition-object, etc. The contextual relationship and syntactic relationship between each topic keyword in the text data to be aggregated are obtained by the above two methods, and weight coefficients can be added to the topic keywords with different relationships respectively. Since the topic keywords in the text data to be aggregated may not appear only once, the higher the frequency of occurrence and the more topic relationships between the topic keywords and other topic keywords, the higher the importance of the topic keywords. Finally, the importance of each topic keyword in a text data to be aggregated can be determined by counting the topic weight values ​​of each topic keyword. By extension, the importance of each topic keyword in each text data to be aggregated can be defined to achieve aggregation of topic data hotly discussed by users and determine hot topics that users are concerned about.

[0068] S14. Summarize the topic weight values ​​of the topic keywords to obtain the total weight value of each topic keyword, and determine the hot topics in the topic keywords according to the total weight value of the topic keywords.

[0069] In this embodiment, the topic weight values ​​of each topic keyword in each group of text data to be aggregated can be added to obtain the total weight value of each topic keyword. The calculated total weight value is used as the basis for determining the popularity of the topic keyword, thereby determining the hot topics among the topic keywords.

[0070] In this embodiment, the topic weight value of the topic keyword can also be adjusted according to the heat value of the text data to be aggregated to which each topic keyword belongs. For example, the text data to be aggregated is sorted according to the heat value. When the text data to be aggregated is ranked first, the topic weight value of the topic keyword corresponding to the text data to be aggregated does not decrease, and is taken as 100%. If the text data to be aggregated is ranked last, the topic weight value of the topic keyword corresponding to the text data to be aggregated is reduced in proportion to the ranking order. Specifically, the topic weight value of the topic keyword can be adjusted in the following manner: A=A1×(R-r+1)÷R, wherein A is the topic weight value of the adjusted topic keyword, A1 is the topic weight value of the topic keyword, r is the ranking order of the hot spot value of the text data to be aggregated corresponding to the topic keyword, and R is the number of text data to be aggregated. In this embodiment, the text data to be aggregated is sorted according to the hot spot value of the text data to be aggregated, and thereby the weight coefficients corresponding to the topic keywords in the text data to be aggregated with different hot spot values ​​are adjusted. Adjustment: When the amount of text data to be aggregated is large, the topic weight value of the topic keyword in the text data to be aggregated with a lower hot spot value can be effectively reduced, so that the total weight value of the topic keyword finally obtained is more in line with reality. When the amount of text data to be aggregated is small, the size of the constant value "1" in the above calculation method can be adjusted. For example, increasing the value of the constant value can reduce the adjustment amount of the topic weight value of the topic keyword, avoiding misjudgment due to the small amount of reference data. Specifically, please refer to the following explanation. For example, when the constant value is "1", if R is 100, the weight coefficient of the topic keyword of the text data to be aggregated ranked first remains unchanged, while the weight coefficient of the topic keyword of the text data to be aggregated ranked last is only 1%. At this time, if the constant "1" is changed to 100, the weight coefficient of the topic keyword of the text data to be aggregated ranked first will double, while the weight coefficient of the topic keyword of the text data to be aggregated ranked last will remain unchanged, and the adjustment of the ranking weight coefficient of the text data to be aggregated will not cause a large gap. In summary, the adjustment can be carried out in the following way: A=A1×(R-r+a)÷R, where A is the topic weight value of the adjusted topic keyword, A1 is the topic weight value of the topic keyword, r is the ranking order of the hotspot values ​​of the text data to be aggregated corresponding to the topic keyword, a is a preset constant, and R is the number of text data to be aggregated.

[0071] In this embodiment, it is not limited to sorting the text data to be aggregated by the hotspot value of the text data to be aggregated, and further adjusting the topic weight values ​​of the topic keywords in the text data to be aggregated. The topic weight values ​​of each topic keyword in the text data to be aggregated can also be adjusted according to the appearance time, text length, reply volume and comment volume of the text data to be aggregated. For example, the longer the appearance time of the text data to be aggregated, the more comments and replies it will have. Therefore, the longer the appearance time of the text data to be aggregated, the more the topic weight values ​​of the topic keywords in the text data to be aggregated should be reduced. The specific adjustment method can refer to the method in the above embodiment. Similarly, the longer the text length, the better the readability of the text data to be aggregated. At this time, the longer the text length, the more the topic weight values ​​of the topic keywords in the text data to be aggregated should be increased. The specific adjustment method can also refer to the method in the above embodiment.

[0072] In this embodiment, this scheme obtains the text data to be aggregated from the interactive text data, and adds topic weight values ​​to the topic keywords according to the topic relationship between the topic keywords in the text data to be aggregated, obtains the total weight value of the topic keywords according to the topic weight value of the topic keywords, and determines the hot topics in the topic keywords according to the total weight value of the topic keywords, thereby realizing the aggregation of topic data hotly discussed by users on the site, so as to quickly determine the hot spots that users on the site are concerned about, and provide data support for the operation work of operators.

[0073] like Figure 2 As shown, in order to ensure that the heat value of the interactive text data is higher than the preset threshold, the step S11 of obtaining the interactive text data from the user interactive text data includes the following steps:

[0074] S21. Obtain popularity data of user interaction text data; the popularity data includes: at least one of the number of comments on the text data, the field to which the text data belongs, and the existence time of the text data.

[0075] In this embodiment, with the rise of the Internet, the interpretation of the term "heat" has also undergone earth-shaking changes. Keyword heat refers to the attention and search degree of users to a certain person or event through search engines and other channels within a certain period of time, and users are also keen to read related content through keyword heat. In daily life, Internet hot words are those words widely used by Internet users. Due to frequent use, the corresponding heat data continues to increase. For example, the frequency of occurrence, the user's search volume, and the user's discussion volume will affect the heat. In this solution, the heat data corresponding to the user interaction text data is obtained, that is, at least one of the text data comment volume, the field to which the text data belongs, and the text data existence time. Of course, it can also be the correlation between the likes and replies and other user interaction text data. For example, words such as digital currency and blockchain do not have any correlation when used alone, but because of the heat generated after their associated use, everyone can also associate them with the effect of their associated use when used alone. Therefore, the correlation between different words can also affect the heat of the data.

[0076] S22: According to a preset scoring standard, determine the popularity data of each user interaction text data and add a corresponding score.

[0077] In this embodiment, corresponding scores are added to the popularity data according to the preset scoring standards. For example, the higher the number of comments on the text data, the higher the corresponding popularity. The longer the data appears, the lower the corresponding popularity of the data. The field to which the text data belongs is the positive and negative comments of users on an event. Although both positive and negative comments will increase the popularity of the topic corresponding to the data, positive comments can promote the popularity of the topic more than negative comments, while negative comments may lead to a decrease in user discussion.

[0078] S23. For each piece of user interaction text data, calculate a heat value according to the score of the heat data of the user interaction text data.

[0079] In this embodiment, the corresponding heat value is calculated according to the score of the heat data of the user interaction text data. For example, a weighted formula based on comments, replies, texts, and timeliness is used. Each score is weighted according to the contribution of the scores of different data to the heat, so as to ensure that the heat value finally calculated is more in line with the actual situation. It is also possible to add heat values ​​to different user interaction text data based on experience, use the score of the heat data of the user interaction text data as input, and use the heat value of the user interaction text data as output to train the heat value model. In this solution, the score of the heat data of different user interaction text data is directly input into the heat value model to directly obtain the corresponding heat value.

[0080] S24: The user interaction text data whose heat value is higher than a preset threshold is regarded as interactive text data.

[0081] In this embodiment, user interaction text data with a heat value higher than a preset threshold is used as user interaction text data to reduce the impact of interference data, improve the accuracy of hot topics determined in this solution, and improve user experience.

[0082] In this embodiment, after determining the hot topics, the information interfaces associated with each hot topic are determined; the hot topics are sorted according to the total weight values ​​corresponding to the hot topics, and the information interfaces corresponding to the hot topics are displayed in sequence.

[0083] In this embodiment, the information interface associated with each hot topic is determined, and the information interface corresponding to the hot topic can be displayed in the form of a list. Since there are multiple hot topics corresponding to the same information interface, this solution can also accumulate the total weight values ​​of the hot topics in the information interfaces corresponding to the multiple hot topics, and then sort the weights of the hot topics corresponding to the information interfaces to determine the display order of each information interface. In this embodiment, the information interface corresponding to the hot topic with a larger total weight value is ranked higher.

[0084] like Figure 3 As shown, an embodiment of the present invention provides a method for determining a hot topic. Figure 3 As shown, the determination method includes the following steps:

[0085] S31. Obtain interactive text data from the user interaction text data; the heat value of the interactive text data is higher than a preset threshold.

[0086] For details of step S31, please refer to the description of step S11, which will not be repeated here in this embodiment.

[0087] S32. Group and aggregate the interactive text data according to the entity names to obtain multiple groups of text data to be aggregated.

[0088] For details of step S32, please refer to the description of step S12, which will not be repeated here in this embodiment.

[0089] S33. Obtain sentence components in the interactive text data.

[0090] In this embodiment, the components of a sentence are called sentence components, also called syntactic components. In a sentence, there is a certain combination relationship between words. According to different relationships, the sentence can be divided into different components. Sentence components are played by words or phrases.

[0091] In this embodiment, a semantic role labeling tool can be used to directly determine entities in interactive text data. Entities can be words such as subject, predicate, object, etc., and then a dependency analysis tool can be used to extract fact triples centered on the predicate, including direct subject, predicate, and object, verb-object relationships with postpositioned attributives, and subject-predicate-verb-complement relationships with prepositional-object relationships. Of course, there are many other sentence components and sentence relationships, which are not listed here in this solution.

[0092] S34. Determine topic keywords in the interactive text data based on sentence components.

[0093] In this embodiment, the topic keywords in this step may be the subject, predicate and object of the interactive text data, while other predicatives, attributives, adverbials and other words that cannot be used to express the importance of the sentence may be used as alternative words.

[0094] In this embodiment, topic keywords in sentence components can also be obtained by training the model. Specifically, sentence components and keywords marked in sentence components can be input into a machine learning model through big data training, and the model can be trained to determine the keywords that can be finally obtained from different sentence components. In this solution, sentence components are directly input into the model to determine topic keywords.

[0095] S35. Obtain the grammatical relationship between topic keywords, configure a first topic weight value for topic keywords with a grammatical relationship, and configure a second topic weight value for topic keywords without a grammatical relationship; the first topic weight value is greater than the second topic weight value.

[0096] In this embodiment, different topic weight values ​​are added to the topic keywords according to the grammatical relationship between different topic keywords. The grammatical relationship in this scheme can be the association between entities, or the sentence relationship between topic keywords. For example, there is a grammatical relationship between two topic keywords with a subject-object relationship, and there is also a grammatical relationship between topic keywords with other sentence relationships.

[0097] In this embodiment, when there is no grammatical relationship between a topic keyword in a sentence and other topic keywords, the topic keyword is an isolated word. Even if such a word is determined to be a topic keyword, its importance is relatively low. The longer an interactive text data is, the more complete the logical description of the interactive text data is, and the more user responses it can get. Determining the popularity of data in this solution is to determine the impact of data interaction on the popularity.

[0098] S36. Summarize the topic weight values ​​of the topic keywords to obtain the total weight value of each topic keyword, and determine the hot topics in the topic keywords based on the total weight value of the topic keywords.

[0099] For details of step S36, please refer to the description of step S14, which will not be repeated here in this embodiment.

[0100] In this embodiment, topic keywords in the interactive text data are determined, and weight coefficients are added to the topic keywords according to grammatical relationships in the topic keywords, so as to quantify the importance of different topic keywords in the interactive text data.

[0101] like Figure 4 As shown, an embodiment of the present invention provides a method for determining a hot topic. Figure 4 As shown, the determination method includes the following steps:

[0102] S41. Obtain interactive text data from user interactive text data; a heat value of the interactive text data is higher than a preset threshold.

[0103] For details of step S41, please refer to the description of step S11, which will not be repeated here in this embodiment.

[0104] S42: Group and aggregate the interactive text data according to the entity names to obtain multiple groups of text data to be aggregated.

[0105] For details of step S42, please refer to the description of step S12, which will not be repeated here in this embodiment.

[0106] S43: Compare the source data of the text data to be aggregated with the preset special symbols to determine whether the source data contains the preset special symbols.

[0107] In this embodiment, the source data of the text data to be aggregated is obtained. Since the source data is data stored in a storage medium, the client or server needs to parse the source data to obtain the corresponding data in order to display the relevant content on the page. In order to place different data in different areas, corresponding preset special symbols need to be added to the data so that when the page is loaded, the corresponding data is loaded into the corresponding area. The preset special symbols can be ##, [], "", etc., or can be a character string named by the user for the data. In this scheme, the source data of the text data to be aggregated is compared with the preset special symbols. When the preset special symbols exist, it means that the source data of this segment contains specially marked data. For example, in a promotional advertisement, it is necessary to emphasize certain words, or in the title of a video, words that need special processing can be used in the corresponding source data to facilitate system recognition. Alternatively, a paragraph can be split and named separately to ensure that different words are processed in different ways to achieve different display effects.

[0108] S44. If the source data has a preset special symbol, the words corresponding to the preset special symbol in the text data to be aggregated are used as candidate keywords.

[0109] In this embodiment, the corresponding words in the text data to be aggregated are determined as candidate keywords according to whether the source data has preset special symbols, so as to reduce the amount of data processing and improve data processing efficiency.

[0110] S45. Determine whether the text data to be aggregated is low-quality data based on the number of repetitions of the candidate keyword in the text data to be aggregated.

[0111] In this embodiment, the higher the number of repetitions of the candidate keywords in the text data to be aggregated, the fewer valid sentences there are in the text data to be aggregated, and the text data to be aggregated is formed by the user constantly repeating a certain paragraph. At this time, if the number of repetitions of the candidate keywords in the text data to be aggregated reaches a preset number of times, the text data to be aggregated is judged to be low-quality data. It can also be judged whether the proportion of the number of repetitions of each candidate keyword to the number of all candidate keywords in the text data to be aggregated is less than a preset proportion. If the proportion of the number of repetitions of any candidate keyword to the number of all candidate keywords in the text data to be aggregated exceeds the preset proportion, the text data to be aggregated is judged to be low-quality data.

[0112] In this embodiment, it is also possible to determine whether the text data to be aggregated is low-quality data by determining the type of the text data to be aggregated or the content length of the text data to be aggregated. For example, some sensitive types or data that cannot be discussed by the public cannot be used as text data to be aggregated, and there are also data generated by the fans of the rice ball. Some TV drama trailers and related interviews cannot be used as text data to be aggregated because they cannot cause extensive discussion. Of course, data with very little substantive content cannot be used as text data to be aggregated. Therefore, in this solution, the category and content length of the text data to be aggregated corresponding to the candidate keywords can also be obtained. When the category of the text data to be aggregated is the same as any preset category or the content length is less than the preset length, the text data to be aggregated is low-quality data, and the preset categories include: blacklist data, rice ball data, and preview data. To determine the category to which the text data to be aggregated belongs, the model obtained by training the machine learning algorithm through samples and the categories to which the samples belong can be identified, and this solution does not make special restrictions on this.

[0113] S46a: When the text data to be aggregated corresponding to the candidate keyword is not low-quality data, the candidate keyword is used as a topic keyword and a third topic weight value is configured.

[0114] S46b: When the text data to be aggregated corresponding to the candidate keyword is low-quality data, the candidate keyword is deleted.

[0115] In this embodiment, it is determined whether the text data to be aggregated corresponding to the candidate keyword is low-quality data. When the text data to be aggregated is low-quality data, the candidate keyword is deleted to reduce the impact of the low-quality data on the final topic keyword, thereby ensuring that the screened candidate keyword better meets the requirements.

[0116] S47, summing up the topic weight values ​​of the topic keywords to obtain the total weight value of each topic keyword, and determining the hot topics in the topic keywords according to the total weight value of the topic keywords.

[0117] For details of step S47, please refer to the description of step S14, which will not be repeated here in this embodiment.

[0118] In this embodiment, the source data corresponding to the text data to be aggregated is screened to determine whether the text data to be aggregated has candidate keywords. After the candidate keywords are determined, it is further determined whether the text data to be aggregated is low-quality data to avoid the topic keywords being words extracted from low-quality data, thereby improving the effectiveness of the topic keywords.

[0119] like Figure 5 As shown, an embodiment of the present invention provides a method for determining a hot topic. Figure 5 As shown, the determination method includes the following steps:

[0120] S51. Obtain interactive text data from user interactive text data; a heat value of the interactive text data is higher than a preset threshold.

[0121] For details of step S51, please refer to the description of step S11, which will not be repeated here in this embodiment.

[0122] S52: Group and aggregate the interactive text data according to the entity names to obtain multiple groups of text data to be aggregated.

[0123] For details of step S52, please refer to the description of step S12, which will not be repeated here in this embodiment.

[0124] S53: Input the text data to be aggregated into a pre-established relation extraction model to obtain all subjects in the text data to be aggregated and the objects corresponding to each subject.

[0125] In this embodiment, a relation extraction model is constructed, and an event subject extraction model can be constructed based on BERT pattern relation learning. The subject in the text data to be aggregated and the object corresponding to each subject are extracted from the text data to be aggregated through the event subject extraction model. Of course, the subject, predicate and object of the sentence can also be directly obtained through the above-mentioned semantic role labeling tool, and the subject of the text data to be aggregated and the object corresponding to each subject in this step can be further obtained.

[0126] S54, inputting the subject and the pre-established grammatical relationship into a pre-established object prediction model to obtain a predicted object.

[0127] In this embodiment, the grammatical relations pre-established in this step are all sentence grammatical relations that can appear in a sentence, such as verb-object relations, subject-predicate-verb-complement relations containing preposition-object relations, etc. Each subject and each pre-established grammatical relation are respectively input into the object prediction model to obtain a predicted object.

[0128] In this embodiment, the object prediction model can be obtained by marking the acquired text data to be aggregated, marking the subject and object in the text data to be aggregated, and then constructing the object prediction model, taking the subject and each pre-established grammatical relationship as the input of the object prediction model, taking the marked object as the output of the object prediction model, and training the parameters of the object prediction model, so that the object prediction model can output the predicted object according to the input subject and the pre-established grammatical relationship. The number of predicted objects under each grammatical relationship can be multiple, which can be determined based on data training. The larger the amount of data, the greater the number of predicted objects that can be obtained in the end. At this time, each predicted object with a possible probability greater than a preset probability is used as the output. The larger the preset probability, the smaller the number of predicted objects obtained in the end, and the smaller the preset probability, the greater the number of predicted objects obtained in the end. In this embodiment, the object prediction model is trained by the text data to be aggregated. The final output result of the object prediction model is the predicted object with a higher probability, that is, the object that exists in a large amount in the text data to be aggregated. If the real object is consistent with or similar to the predicted object, it means that the subject and object in the text data to be aggregated are data that are widely commented on or widely used by users, which further verifies whether the subject and object in the text data to be aggregated can be used as topic keywords.

[0129] S55. Determine whether the predicted object is similar to the object corresponding to the subject.

[0130] In this embodiment, since the input in this scheme is the subject and the preset grammatical relationship, the final predicted object will be quite different from the object in the actual situation. In this embodiment, to judge whether the predicted object is similar to the real object, the similarity can be judged by directly comparing whether the predicted object is consistent with the real object, or the similarity can be judged from the grammatical level. For example, if the predicted object and the object corresponding to the subject have the same part of speech, the predicted object is considered to be similar to the real object. It can also be judged whether the predicted object and the object corresponding to the subject have the same meaning, such as house and house. If the predicted object and the object corresponding to the subject have the same meaning, the predicted object is considered to be similar to the real object, so as to improve the adaptability of this scheme.

[0131] S56a: If the predicted object is similar to the object corresponding to the subject, the subject and the object corresponding to the subject are used as topic keywords, and a fourth topic weight value is configured for the topic keyword.

[0132] S56b. If the predicted object is not similar to the object corresponding to the subject, the subject and the object corresponding to the subject are used as topic keywords, and a fifth topic weight value is configured for the topic keywords; wherein the fourth topic weight value is greater than the fifth topic weight value.

[0133] In this embodiment, if the predicted object is similar to the object corresponding to the subject, it is considered that there is a grammatical relationship between the subject and the object, and the subject and the object are used as topic keywords, and a fourth topic weight value is added to the topic keywords. Otherwise, there is no grammatical relationship between the subject and the object, and no other processing is performed.

[0134] S57, summing up the topic weight values ​​of the topic keywords to obtain the total weight value of each topic keyword, and determining the hot topics in the topic keywords according to the total weight value of the topic keywords.

[0135] For details of step S57, please refer to the description of step S14, which will not be repeated here in this embodiment.

[0136] like Figure 6 As shown, in this embodiment, in S53, the text data to be aggregated is input into a pre-established relation extraction model to obtain all subjects in the text data to be aggregated and the objects corresponding to each subject, including the following steps:

[0137] S61. Encode each character in the text data to be aggregated according to a preset correspondence between characters and numbers to obtain a digital sequence that is the same as the order of the characters in the text data to be aggregated.

[0138] In this embodiment, since the system cannot directly understand the meaning of the text when training the relationship extraction model, in this solution, when training the relationship extraction model, the text can be encoded and input into the relationship extraction model, and the output numbers can be converted into corresponding text when outputting, which can improve the efficiency of training.

[0139] S62: Input the digital sequence into the relational extraction model to obtain all subjects in the text data to be aggregated and the objects corresponding to each subject.

[0140] In this embodiment, the model training process is carried out by converting text into corresponding digital sequences, thereby reducing the time consumption of the system in identifying and comparing data, and improving the data processing efficiency and training efficiency of the model. In this solution, when using the relation extraction model, the corresponding input is converted into a digital input relation extraction model, and the output numbers are converted into corresponding text at the final output, so that the subject in the text data to be aggregated and the object corresponding to each subject can be obtained.

[0141] In this embodiment, the object prediction model obtained through training predicts the object of the subject in the text data to be aggregated under different grammatical relationships, and obtains the predicted object of the subject under different grammatical relationships. Since the data used in training the object prediction model are the text data to be aggregated, the object predicted by the object prediction model is the object with a higher frequency of occurrence corresponding to the subject in the text data to be aggregated. Therefore, this scheme can quickly determine the subjects and objects with a higher frequency of occurrence in the text data to be aggregated, thereby improving the screening efficiency of this scheme for topic keywords.

[0142] In a specific embodiment, training samples of the relation extraction model can be generated by marking the subjects, predicates, and objects in a large number of common sentences. Furthermore, in this solution, the text of the sentences in the marked training samples can be converted into numbers, the converted numbers can be used as the input of the relation extraction model, and the various tags in each sentence can be used as the output of the relation extraction model to train the above-mentioned relation extraction model.

[0143] When using the relational extraction model, the text data to be aggregated is converted into numbers and input into the relational extraction model, so as to obtain the subject, predicate and object corresponding to the aggregated text data output by the relational extraction model.

[0144] like Figure 7 As shown, an embodiment of the present invention provides a hot topic determination system, and the determination system includes: a first acquisition unit 11, a second acquisition unit 12, a weight determination unit 13 and a determination unit 14.

[0145] In this embodiment, the first acquisition unit 11 is used to acquire interactive text data from user interactive text data; the heat value of the interactive text data is higher than a preset threshold.

[0146] In this embodiment, the second acquisition unit 12 is used to group and aggregate the interactive text data according to the entity name to obtain multiple groups of text data to be aggregated.

[0147] In this embodiment, the weight determination unit 13 is used to extract topic keywords from the text data to be aggregated for each group of text data to be aggregated, and determine the topic weight value of each topic keyword according to the topic relationship between the topic keywords.

[0148] In this embodiment, the determination unit 14 is used to aggregate the topic weight values ​​of the topic keywords to obtain the total weight value of each topic keyword, and determine the hot topics in the topic keywords according to the total weight value of the topic keywords.

[0149] In this embodiment, the first acquisition unit 11 is specifically used to obtain the popularity data of user interaction text data; the popularity data includes: at least one of the number of comments on the text data, the field to which the text data belongs, and the existence time of the text data; according to a preset scoring standard, the score corresponding to the popularity data of each user interaction text data is determined; for each user interaction text data, the popularity value is calculated according to the score of the popularity data of the user interaction text data; the user interaction text data with a popularity value higher than a preset threshold is regarded as interactive text data.

[0150] In this embodiment, the second acquisition unit 12 is specifically configured to acquire sentence components in the text data to be aggregated; and determine topic keywords in the text data to be aggregated according to the sentence components.

[0151] In this embodiment, the weight determination unit 13 is specifically used to obtain the grammatical relationship between topic keywords, configure a first topic weight value for topic keywords with a grammatical relationship, and configure a second topic weight value for topic keywords without a grammatical relationship; the first topic weight value is greater than the second topic weight value.

[0152] Alternatively, the weight determination unit 13 is specifically used to compare the source data of the text data to be aggregated with the preset special symbols to determine whether the source data contains the preset special symbols; if the source data contains the preset special symbols, the words corresponding to the preset special symbols in the text data to be aggregated are used as candidate keywords, and the candidate keywords are used as topic keywords to configure the third topic weight value.

[0153] In this embodiment, the determination system also includes: a judgment unit, which is used to judge whether the text data to be aggregated is low-quality data based on the number of repetitions of the candidate keyword in the text data to be aggregated; when the text data to be aggregated corresponding to the candidate keyword is not low-quality data, the candidate keyword is used as a topic keyword and a third topic weight value is configured; when the text data to be aggregated corresponding to the candidate keyword is low-quality data, the candidate keyword is deleted.

[0154] Alternatively, the weight determination unit 13 is specifically used to input the text data to be aggregated into a pre-established relationship extraction model to obtain all subjects in the text data to be aggregated and the objects corresponding to each subject; input the subjects and the pre-established grammatical relations into a pre-established object prediction model to obtain predicted objects; determine whether the predicted object is similar to the object corresponding to the subject; if the predicted object is similar to the object corresponding to the subject, then use the subject and the object corresponding to the subject as topic keywords, and configure a fourth topic weight value for the topic keywords; if the predicted object is not similar to the object corresponding to the subject, then use the subject and the object corresponding to the subject as topic keywords, and configure a fifth topic weight value for the topic keywords; wherein the fourth topic weight value is greater than the fifth topic weight value.

[0155] In this embodiment, the second acquisition unit 12 is specifically used to encode each character in the text data to be aggregated according to a preset correspondence between characters and numbers, and obtain a digital sequence with the same order as the characters in the text data to be aggregated; the digital sequence is input into the relational extraction model to obtain all subjects in the text data to be aggregated and the objects corresponding to each subject.

[0156] The determination system further comprises: a display unit for determining an information interface associated with each hot topic;

[0157] Hot topics are sorted according to their total weight values, and the information interfaces corresponding to each hot topic are displayed in sequence.

[0158] The functions of the device in the embodiment of the present invention have been described in the above method embodiment, so for details not provided in the description of this embodiment, please refer to the relevant description in the above embodiment, and no further description will be given here.

[0159] like Figure 8 As shown, an embodiment of the present invention provides an electronic device, including a processor 1110, a communication interface 1120, a memory 1130 and a communication bus 1140, wherein the processor 1110, the communication interface 1120, and the memory 1130 communicate with each other through the communication bus 1140;

[0160] Memory 1130, used for storing computer programs;

[0161] The processor 1110 is configured to implement the hot topic determination method in any of the above embodiments when executing the program stored in the memory 1130 .

[0162] In the electronic device provided by the embodiment of the present invention, the processor 1110 obtains interactive text data from user interaction text data by executing the program stored in the memory 1130, groups and aggregates the interactive text data according to the entity name, and obtains multiple groups of text data to be aggregated, and adds topic weight values ​​to the topic keywords according to the topic relationship between the topic keywords in the text data to be aggregated, obtains the total weight value of the topic keywords according to the topic weight values ​​of the topic keywords, and determines the hot topics in the topic keywords according to the total weight values ​​of the topic keywords, thereby realizing the aggregation of topic data hotly discussed by users on the site, so as to quickly determine the hot spots that users on the site are concerned about, and provide data support for the operation work of operators.

[0163] The communication bus 1140 mentioned in the above electronic device can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus 1140 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0164] The communication interface 1120 is used for communication between the above electronic device and other devices.

[0165] The memory 1130 may include a random access memory 1130 (RAM), or may include a non-volatile memory 1130 (non-volatile memory), such as at least one disk memory 1130. Optionally, the memory 1130 may also be at least one storage device located away from the processor 1110.

[0166] The above-mentioned processor 1110 can be a general-purpose processor 1110, including a central processing unit 1110 (CPU), a network processor 1110 (NP), etc.; it can also be a digital signal processor 1110 (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.

[0167] An embodiment of the present invention provides a computer-readable storage medium, which stores one or more programs. The one or more programs can be executed by one or more processors 1110 to implement the method of any of the above embodiments.

[0168] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function according to the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, a computer, a server or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or a data center that includes one or more available media integrated. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state hard disk SolidStateDisk (SSD)), etc.

[0169] 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 determining hot topics, characterized in that: Methods for determination include: Acquire interactive text data from user interactive text data; the heat value of the interactive text data is higher than a preset threshold; Grouping and aggregating the interactive text data according to entity names to obtain multiple groups of text data to be aggregated; For each group of text data to be aggregated, extract topic keywords in the text data to be aggregated, and determine a topic weight value of each topic keyword according to the topic relationship between the topic keywords; The topic weight values ​​of the topic keywords are aggregated to obtain a total weight value of each topic keyword, and the hot topics in the topic keywords are determined based on the total weight value of the topic keywords.

2. The method for determining hot topics according to claim 1, characterized in that: The step of obtaining interactive text data from user interactive text data includes: Acquire popularity data of user interaction text data; the popularity data includes: at least one of the number of comments on the text data, the field to which the text data belongs, and the existence time of the text data; Determine the score corresponding to the popularity data of each piece of the user interaction text data according to a preset scoring standard; For each piece of the user interaction text data, calculating the heat value according to the score of the heat data of the user interaction text data; The user interaction text data whose heat value is higher than a preset threshold is used as interactive text data.

3. The method for determining hot topics according to claim 1, characterized in that: The extracting of topic keywords from the text data to be aggregated and determining the topic weight value of each topic keyword according to the topic relationship between the topic keywords includes: Acquire sentence components in the text data to be aggregated; Determining topic keywords in the text data to be aggregated according to the sentence components; The grammatical relationship between the topic keywords is obtained, a first topic weight value is configured for the topic keywords with a grammatical relationship, and a second topic weight value is configured for the topic keywords without a grammatical relationship; the first topic weight value is greater than the second topic weight value.

4. The method for determining hot topics according to claim 1, characterized in that: The extracting of topic keywords from the text data to be aggregated and determining the topic weight value of each topic keyword according to the topic relationship between the topic keywords includes: Comparing the source data of the text data to be aggregated with the preset special symbol to determine whether the source data contains the preset special symbol; If the source data has a preset special symbol, the words corresponding to the preset special symbol in the text data to be aggregated are used as candidate keywords, and the candidate keywords are used as the topic keywords, and a third topic weight value is configured.

5. The method for determining hot topics according to claim 4, characterized in that: Before using the candidate keyword as the topic keyword and configuring the third topic weight value, the method further includes: According to the number of repetitions of the candidate keyword in the text data to be aggregated, determining whether the text data to be aggregated is low-quality data; When the text data to be aggregated corresponding to the candidate keyword is not low-quality data, the candidate keyword is used as the topic keyword, and a third topic weight value is configured; When the text data to be aggregated corresponding to the candidate keyword is low-quality data, the candidate keyword is deleted.

6. The method for determining hot topics according to claim 1, characterized in that: The extracting of topic keywords from the text data to be aggregated and determining the topic weight value of each topic keyword according to the topic relationship between the topic keywords includes: Inputting the text data to be aggregated into a pre-established relation extraction model to obtain all subjects in the text data to be aggregated and objects corresponding to each subject; Inputting the subject and the pre-established grammatical relationship into a pre-established object prediction model to obtain a predicted object; Determining whether the predicted object is similar to an object corresponding to the subject; If the predicted object is similar to the object corresponding to the subject, the subject and the object corresponding to the subject are used as topic keywords, and a fourth topic weight value is configured for the topic keyword; If the predicted object is not similar to the object corresponding to the subject, the subject and the object corresponding to the subject are used as topic keywords, and a fifth topic weight value is configured for the topic keyword; The fourth topic weight value is greater than the fifth topic weight value.

7. The method for determining hot topics according to claim 6, characterized in that: The step of inputting the text data to be aggregated into a pre-established relation extraction model to obtain all subjects in the text data to be aggregated and objects corresponding to each subject includes: According to a preset correspondence between characters and numbers, each character in the text data to be aggregated is encoded to obtain a digital sequence that is the same as the order of the characters in the text data to be aggregated; The digital sequence is input into the relation extraction model to obtain all subjects in the text data to be aggregated and the objects corresponding to each subject.

8. The method for determining hot topics according to any one of claims 1 to 7, characterized in that: The determination method further comprises: Determine the information interface associated with each hot topic; The hot topics are sorted according to the total weight values ​​corresponding to the hot topics, and the information interfaces corresponding to the hot topics are displayed in sequence.

9. A hot topic determination system, characterized in that: The determination system comprises: A first acquisition unit is used to acquire interactive text data from user interactive text data; the heat value of the interactive text data is higher than a preset threshold; A second acquisition unit, configured to group and aggregate the interactive text data according to entity names to obtain multiple groups of text data to be aggregated; A weight determination unit, for extracting topic keywords from each group of text data to be aggregated, and determining a topic weight value of each topic keyword according to the topic relationship between the topic keywords; The determination unit is used to summarize the topic weight values ​​of each topic keyword to obtain the total weight value of each topic keyword, and determine the hot topics in the topic keywords according to the total weight value of the topic keywords.

10. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; Memory, used to store computer programs; The processor is used to implement the hot topic determination method described in any one of claims 1 to 8 when executing the program stored in the memory.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the hot topic determination method described in any one of claims 1 to 8.

Images