Hotspot information monitoring and processing system and method

Abstract

The invention relates to the technical field of big data processing, in particular to a hotspot information monitoring and processing system and method.The hotspot information monitoring and processing system comprises a hotspot information monitoring module, a hotspot information collecting module, a hotspot information processing module and a put content processing module; the hot spot information monitoring module is used for monitoring hot spot information of the current Internet, the hot spot information acquisition module is used for acquiring the monitored hot spot information, and the hot spot information processing module is used for processing the acquired hot spot information and generating a processing result; and the delivery content processing module processes the media content of the enterprise in combination with the processing result to generate new media delivery content. According to the method, the current hotspot and the media content of the enterprise can be fused and then put, so that the quantity and efficiency of obtaining clue information by the enterprise are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of big data processing, in particular to a hot information monitoring processing system and method. BACKGROUND

[0002] With the rapid development of the Internet, big data and self-media, hot information on various network platforms emerges in endlessly. The so-called hot information usually refers to public information with high public attention, high click volume and / or high evaluation in a period of time. In the prior art, there are systems for monitoring and classifying these hot information. However, if the enterprise's own media content is integrated with the current hot information based on the heat of the hot information, the quantity of the information obtained by the enterprise from the customer group will undoubtedly be greatly increased. However, there is currently no related technology reported for integrating the hot information with the enterprise's media content. SUMMARY

[0003] The first aspect of the present application is to provide a hot information monitoring processing system to integrate the current hot information with the enterprise's own media content before putting it into the market, so as to improve the quantity and efficiency of the information obtained by the enterprise.

[0004] The hot information monitoring processing system comprises a hot information monitoring module, a hot information acquisition module, a hot information processing module and a putting content processing module. The hot information monitoring module is used for monitoring the current hot information on the Internet. The hot information acquisition module is used for acquiring the monitored hot information. The hot information processing module is used for processing the acquired hot information and generating a processing result. The putting content processing module processes the enterprise's own media content in combination with the processing result to generate new media putting content. The system further comprises a dynamic putting strategy module. The dynamic putting strategy module is used for automatically associating hot tags with customer industry attributes to generate a putting matrix. The putting matrix comprises a hot dimension, a customer dimension and a putting action. When the heat value of the applied hot information exceeds a set heat value and the industry matching degree exceeds a set correlation value, the putting action is to generate new media putting content and put it into the Internet platform. When the click volume of the new media content put in a period of time exceeds a set rate value and the historical customer click volume is greater than a set click volume value, the putting action is to add a putting budget. When the participating customer group is a specific age group in a period of time and the age group coincides with the target customer group, the Internet platform matching the specific age group is put into focus.

[0005] Explanation: The hot information described in the present application includes hot videos and hot searches on various social platforms.

[0006] The clue information described in the present application refers to the information of potential customers obtained by the enterprise through any channel.

[0007] The beneficial effects of the present application are that: the present application can monitor the hot information of each platform in the Internet in real time through the hot information monitoring module, when a certain hot information is monitored, the hot information is collected through the hot information collection module, so as to realize the mining and utilization of the hot information, the present application further processes the collected hot information through the hot information processing module and generates a processing result, finally, the content processing module combines the processing result to process the media content of the enterprise itself, generates new media content and puts it. Through the above mode, the hot information is monitored to understand which is the hot content in real time, based on the hot content, the related hot media file content of the enterprise industry is edited and improved to increase the click volume and activity of the enterprise, such as editing and generating the media file carrying the hot word or hot information related slice content, so that in addition to the enterprise itself putting its own media information on each platform, through the use of the system, the current hot spot can also be linked, the click volume, attention and activity of the media content put by the enterprise are increased, the channel for obtaining the clue information is increased, and the efficiency for obtaining the clue information is greatly improved through the use of the system.

[0008] The present application can generate a multi-dimensional delivery matrix according to the hot tag and the customer industry attribute through the dynamic delivery strategy module, the present application flexibly sets and adjusts the delivery action according to the hot dimension and the customer dimension, so that the delivery can be more reasonable, the target customer group can be more accurately locked, and more clue information can be obtained.

[0009] The preferred embodiment of the present application is that: the hot information processing module further comprises an emotional tendency analysis unit, an industry relevance matching unit and a hotness value calculation unit, the emotional tendency analysis unit is used for analyzing the emotional tendency of the currently collected hot information, and screening out the hot information with positive emotion and emotional value, the industry relevance matching unit is used for screening out the hot information related to the industry and calculating the matching degree, and the hot information hotness value calculation unit is used for calculating the hotness value of the hot information in the historical time period, the hot information processing module grades the hot information based on the hotness value, the industry matching degree and the emotional value of the hot information.

[0010] The beneficial effect is that: in the application, all collected hot information is screened and graded from three dimensions by the hot information processing module, first, the mood tendency analysis unit analyzes the mood tendency of the current collected hot information, and screens out hot information with positive emotion, and gives emotional value to the hot information with positive emotion according to the positive degree, second, the industry relevance matching unit screens out hot information related to the industry and calculates the matching degree, so that the screened hot information is related to the industry, and the fusion with the media content of the enterprise itself is more natural and not in conflict, and third, the hot information hotness value calculation unit calculates the hotness value of the hot information in the historical time period, so as to accurately understand the hotness of each hot information, and the hot information processing module grades the hot information based on the hotness value, industry matching degree and emotional value of the hot information, so as to screen out hot information with better comprehensive situation and more suitable for the enterprise for the enterprise to use.

[0011] The preferred embodiment of the application is that: the hot information processing module further comprises a hot semantic analysis unit and a core word extraction unit, the hot semantic analysis unit is used for semantic analysis of the screened hot information, and the core word extraction unit is used for extracting the core word after semantic analysis.

[0012] The beneficial effect is that: the application screens out the hot information by the hot semantic analysis unit, so that the content of the hot information can be deeply analyzed, and the essence and hidden meaning can be understood, and the core word extraction unit can extract the core word after semantic analysis, so that the hot information can be simplified and accurately used.

[0013] The preferred embodiment of the application is that: the delivery content processing module comprises a content reorganization unit and a hot visual element adding unit, the content reorganization unit is used for inserting the extracted core word into the media content of the enterprise itself, and the hot visual element adding unit is used for adding visual elements related to the current hot spot to the media content of the enterprise itself, and generating new media delivery content.

[0014] The beneficial effect is that: the application can usually adopt two or three ways for the utilization of hot information, of course, not limited to these ways, including inserting the extracted core word into the enterprise's own media content through the content reorganization unit, or adding visual elements related to the current hot spot to the enterprise's own media content through the hot spot visual element adding unit, or inserting the core word into the enterprise's own media content and adding the visual elements related to the current hot spot to the enterprise's own media content, and then putting the generated new media content. Since the hot spot word and / or hot spot element are added to the enterprise's own media content, the new media content of the enterprise based on the current hot spot will increase the click volume, attention, comments and activity compared with the media file without adding any hot spot content, thereby increasing more lead information for the enterprise.

[0015] The preferred embodiment of the application is that: the hot spot information processing module includes a hot spot information slicing unit, which is used to slice the received hot spot information and identify and extract key hot spot segments, and the delivery content processing module is used to implant the enterprise's own media content into the key hot spot segments to generate new media delivery content, or the delivery content processing module extracts segments or special effects from the key hot spot segments and applies them to the enterprise's own media content to generate new media delivery content.

[0016] The beneficial effect is that: as another way to integrate hot spots with the enterprise's media content, the application can also slice the received hot spot information through the hot spot information slicing unit and identify and extract key hot spot segments in each segment, and then implant the enterprise's own media content into the key hot spot segments to generate new media delivery content, so that when other people click, pay attention to or evaluate the hot spot of the segment, the enterprise's relevant products, introductions or services can be displayed to the visitors, or segments or special effects are extracted from the key hot spot segments and applied to the enterprise's own media content, so as to attract more visitors through the hot spot segments or special effects, thereby increasing the lead information obtained by the enterprise through the above-mentioned ways.

[0017] The preferred embodiment of the application is that: the analysis method of the industry relevance matching unit is as follows: Construct an industry keyword library; Calculate the cosine similarity of the hot spot text and the keyword library; Matching degree calculation formula: In the formula, ki is the i-th industry keyword, T is the hot spot text, TF-IDF(ki) is the TF-IDF weight of the keyword ki, and Sim(ki, T) is the word vector similarity between the keyword ki and the hot spot text T.

[0018] A preferred embodiment of the present invention further includes a testing module, which is used to automatically monitor the original media content and the newly deployed media content that incorporates trending topics, and to compare and analyze the click-through rate and lead conversion rate of the two.

[0019] The beneficial effects are as follows: This invention uses a testing module to monitor the original media content and the newly launched media content that incorporates trending topics, and further analyzes the click-through rate and lead conversion rate of both. This allows for an accurate understanding of the impact of media content without and with trending topics on the company's lead generation, as well as the differences in lead generation brought by media content that incorporates different trending topics. This provides strong support for companies to predict more suitable trending topics and content that can increase lead generation, and optimizes the integration of trending topics and company media content in subsequent integration processes.

[0020] A preferred embodiment of the present invention is as follows: the popularity value calculation unit calculates the popularity value within a historical time period based on a time decay model, combining click volume, comment volume, and repost volume indicators, as shown in the following formula: Where: R represents the popularity value, C represents the number of clicks, R f λ represents the number of reposts, M represents the number of comments; Wc, Wr, and Wm are the weights of clicks, reposts, and comments, respectively, reflecting the contribution of each indicator to the popularity; λ is the decay factor, which controls the rate at which the popularity decays over time; t is the time interval, calculated from the moment the hot information was published.

[0021] This invention provides another preferred method for calculating popularity value: the popularity value calculation unit is based on a time decay model, and comprehensively considers click volume and forwarding volume indicators to calculate the popularity value within a historical time period, as shown in the following formula: Among them, H t C represents the heat value. i For clicks, R i For reposts, w c w r λ is the weight, and λ is the decay factor.

[0022] The second aspect of this invention is to provide a hotspot information monitoring and processing method that can improve the quantity and efficiency of enterprises in obtaining customer leads.

[0023] The methods for monitoring and processing trending information include the following: Monitor current trending information on the internet; Collect and monitor hotspot information; analyzing the emotional tendency of the current collected hot information, screening out the hot information with positive emotion and emotional value, screening out the hot information related to the industry and calculating the matching degree, calculating the heat value of the hot information in the historical time period, and classifying the hot information based on the heat value, industry matching degree and emotional value of the hot information; performing semantic analysis and core word extraction on the hot information with a grade higher than the preset grade; inserting the extracted core word into the enterprise's own media content, adding visual elements related to the current hot spot to the enterprise's own media content, and generating new media delivery content; or slicing the received hot information, identifying and extracting key hot spot segments, implanting the enterprise's own media content into the key hot spot segments to generate new media delivery content, or extracting segments or special effects from the key hot spot segments and applying them to the enterprise's own media content to generate new media delivery content; The dynamic delivery strategy module automatically associates hot spot tags with customer industry attributes, generates a delivery matrix, and the delivery matrix includes hot spot dimensions, customer dimensions and delivery actions. When the heat value of the applied hot information exceeds the set heat value, and the industry matching degree exceeds the set value of the correlation degree, the delivery action is to generate new media delivery content and deliver it on the Internet platform. When the click-through rate of the new media content delivered in a period of time exceeds the set rate value, and the historical customer click-through rate is greater than the set value, the delivery action is to add delivery budget. When it is monitored that the participating customer group is a specific age group in a period of time, and the age group coincides with the target customer group, the Internet platform matching the specific age group is delivered. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 The schematic framework diagram of the hot information monitoring and processing system of the present application. DETAILED DESCRIPTION

[0025] The preferred embodiments of the present application are described below in conjunction with the accompanying drawings, and it should be understood that the preferred embodiments described below are only used to explain and describe the present application, and do not limit the scope of protection of the present application.

[0026] The present application is further described in detail below through the preferred specific embodiments: As shown in the accompanying drawings, Figure 1 The hot information monitoring and processing system disclosed in the present embodiment is applied to the scenario of fusing the hot information obtained by the enterprise with its own media content, and delivering the fused media content to obtain the customer group clue information of the enterprise based on the heat of the hot spot, so as to improve the number and efficiency of obtaining customer group clue information.

[0027] The hotspot information monitoring processing system of the embodiment includes a hotspot information monitoring module, a hotspot information collection module, a hotspot information processing module, and a delivery content processing module. The hotspot information monitoring module is configured to monitor the hotspot information of the current Internet. In the embodiment, the Internet can refer to various platforms such as Douyin, Xiaohongshu, and Weibo, and the hotspot information can be a hotspot video, a hot search word, or a hotspot text content. The monitoring module can preliminarily screen out the hotspot information by monitoring the click volume, comment volume, and / or attention volume of the information published on various platforms.

[0028] The hotspot information collection module is configured to collect the monitored hotspot information. In the embodiment, the hotspot information collection module can collect the hotspot information by using a crawler technology.

[0029] The hotspot information processing module is configured to process the collected hotspot information and generate a processing result. The delivery content processing module processes the media content of an enterprise itself in combination with the processing result to generate new media delivery content.

[0030] Specifically, the hotspot information processing module in the embodiment further includes a hotspot semantic analysis unit and a core word extraction unit. The hotspot semantic analysis unit is configured to perform semantic analysis on the screened hotspot information, and the core word extraction unit is configured to extract the core word after semantic analysis.

[0031] The semantic analysis unit uses a natural language processing (NLP) technology to perform deep semantic analysis on the text content of the hotspot information. Specifically, the semantic analysis includes: Entity recognition: identifying key entities such as names, places, organizations, and events in the hotspot information.

[0032] Sentiment analysis: determining the sentiment polarity (positive, negative, or neutral) and sentiment intensity of the text.

[0033] Topic modeling: extracting the topic distribution of the hotspot information by using an LDA (Latent Dirichlet Allocation) algorithm.

[0034] Topic modeling is used to automatically discover high-frequency discussion topics and hotspot issues in business interactions. The following methods are used: Algorithm selection: An unsupervised learning algorithm such as LDA (Latent Dirichlet Allocation) or NMF (Non-Negative Matrix Factorization) is used to extract latent topics from text data.

[0035] Data preprocessing: The collected text is cleaned (remove irrelevant characters, emoticons, etc.), and tokenized and stopword-removed.

[0036] Build a dictionary (Dictionary) and document, term matrix (Document-Term Matrix) as model input.

[0037] Model training and topic extraction: Set the number of topics (K value), and select the optimal number of topics by perplexity (Perplexity) or coherence score (Coherence Score).

[0038] Use LDA or NMF algorithm to decompose the document, term matrix, to get the term distribution of each topic and the topic distribution of each document.

[0039] For each topic, extract its keywords and manually induce topic labels.

[0040] Topic evolution analysis: Slice the data by time window (e.g. weekly or monthly), model the topics separately, analyze the evolution trend of topic intensity and keywords over time, and identify emerging or declining topics.

[0041] Output results.

[0042] Semantic role labeling: analyze the relationship between predicates and entities in a sentence, such as "who did what to whom".

[0043] Contextual association analysis: combine the source, comments and forwarding content of hot information to analyze its implicit social context and transmission intention.

[0044] For example, for the hot text "a certain brand of new energy vehicle breaks through 1000 kilometers of endurance", semantic analysis can output: Entities: {Brand: X, Product: New Energy Vehicle, Technical Indicator: Endurance 1000 kilometers} Sentiment: Positive (sentiment value 0.8) Topic: Technological breakthrough, new energy vehicle Semantic roles: {Subject: A certain brand, Action: Breakthrough, Object: Endurance Mileage} After the hot information monitoring module preliminarily monitors all the hot information of the current period on various platforms, the hot information processing module is used for screening. Specifically, the hot information processing module further includes an emotional tendency analysis unit, an industry relevance matching unit, and a hotness value calculation unit.

[0045] Emotional tendency analysis method: use a deep learning-based sentiment classification model to classify hot text sentiment and output sentiment probability value (0-1).

[0046] The sentiment analysis unit adopts a deep learning-based sentiment classification model, specifically using a pre-trained BERT (Bidirectional Encoder Representations from Transformers) model for fine-tuning to adapt to the sentiment analysis task of hot texts. The following are the detailed implementation steps: Model architecture and training: Model selection: Use the BERT-base Chinese pre-trained model, add a fully connected layer and activation function to its output layer, to output the probability distribution of positive, negative or neutral sentiment.

[0047] Training data: Use public Chinese sentiment analysis datasets (such as ChnSentiCorp, NLPCC sentiment analysis dataset) for fine-tuning. Data preprocessing includes text cleaning, word segmentation and serialization.

[0048] Training parameters: Use cross-entropy loss function, Adam optimizer, learning rate set to 2e-5, batch size 32, training period 3-5 rounds.

[0049] Convert sentiment probability value to sentiment value: The model output is the probability value of positive sentiment, which is converted to a sentiment value Q between 0 and 100 through linear scaling.

[0050] Conversion formula: Q = round(Ppositive × 100), where round() represents rounding to the nearest integer.

[0051] Example: When the input hot text is "This new phone takes amazing photos!", the model outputs a positive probability of 0.92, so the sentiment value Q = 92.

[0052] Sentiment screening: Set a sentiment threshold (such as positive sentiment probability ≥ 0.6), and only keep hot information with positive emotions. The sentiment value Q is used for subsequent grade division, and the higher the value, the more positive the sentiment.

[0053] Calculation example: Input hot text: "This new phone takes amazing photos!" Model output: Positive probability 0.92 → Sentiment value Q = 92 (normalized to 0-100).

[0054] Analysis method of industry relevance matching unit: Build an industry keyword library, such as "range" "charge" "autonomous driving" for the automotive industry; Calculate the cosine similarity between hot texts and keyword library, combined with BERT word vector representation; Matching degree calculation formula: In the formula, ki is the i-th industry keyword, T is the hot text, TF-IDF(ki) is the TF-IDF weight of keyword ki, and Sim(ki, T) is the word vector similarity between keyword ki and hot text T. TF (Term Frequency): measures the frequency of a word appearing in a document, IDF (Inverse Document Frequency): measures the general importance of a word, common words have low IDF, rare words have high IDF, TF-IDF = TF x IDF, used to evaluate the importance of a word to a document. Formula meaning: usually use word vector models (such as Word2Vec, BERT, etc.) to convert words and texts into vectors. For each keyword, multiply its TF-IDF weight by its similarity to the hot text, and then sum them up. Equivalent to the sum of weighted similarities, with weights being the importance of keywords (TF-IDF).

[0055] Sim(ki, T) represents the cosine similarity or other similarity measures between the vector of keyword ki and the vector of hot text T, the closer the value is to 1, the more similar the semantics.

[0056] Calculation example: hot text: "a certain brand of new energy vehicle breaks through 1000 kilometers of endurance", matching degree with the keyword library of the automobile industry is calculated as 85% (threshold can be set to 80%).

[0057] In this embodiment, the heat value calculation unit calculates the heat value in the historical time period (such as 24 hours) based on the time decay model, integrating the click volume, comment volume, and forwarding volume indicators. The formula is as follows: Where: R represents the heat value, C represents the click volume, Rf represents the forwarding volume, and M represents the comment volume; Wc, Wr, and Wm are the weights of click volume, forwarding volume, and comment volume, respectively, reflecting the contribution of each indicator to heat; λ is the decay factor, controlling the rate of heat decay over time; t is the time interval (unit: hours), calculated from the time of publishing the hot information.

[0058] Weight calculation method: The weight values are determined through historical data regression analysis or A / B testing to ensure that the weights reasonably reflect the importance of the indicators. For example: Wc=0.5 (click volume weight, assuming that click volume is the most important), Wr=0.3 (forwarding volume weight), Wm=0.2 (comment volume weight), and the weights must satisfy the normalization condition: Wc+Wr+Wm=1.

[0059] Decay factor calculation method: the decay factor λ is set based on the average life cycle of the hot spot information. For example, for a hot spot period of 24 hours, set the half-life period to 12 hours, then: In practical applications, λ can be adjusted according to business needs: if the hot spot decays faster, increase λ; otherwise, decrease.

[0060] Calculation example: a certain hot spot has 100,000 clicks and 5,000 forwards in the last 24 hours. The heat value R is calculated to be 75 (normalized to 0-100) by substituting the formula.

[0061] In another embodiment, the heat value calculation unit calculates the heat value in the historical time period based on the time decay model, and integrates the click volume and forwarding volume indicators. The formula is as follows: Where H t is the heat value, C i is the click volume, R i is the forwarding volume, w c and w r are weights, and λ is the decay factor.

[0062] In another embodiment, the weight calculation method and the decay factor calculation method related to the heat value are the same as in this embodiment.

[0063] The sentiment tendency analysis unit is used to analyze the sentiment tendency of the currently collected hot spot information, and to filter out hot spot information with positive sentiment and emotional values. In this embodiment, hot spot information with negative or suspected negative energy is filtered out, and hot spot information with positive sentiment is given different emotional values according to different degrees.

[0064] The industry relevance matching unit is used to filter out hot spot information related to the industry and calculate the matching degree. The hot spot information heat value calculation unit is used to calculate the heat value of the hot spot information in the historical time period. The hot spot information processing module divides the hot spot information into grades based on the heat value, industry matching degree and emotional value of the hot spot information.

[0065] In this embodiment, the sentiment tendency analysis unit identifies the sentiment tendency and assigns emotional values by analyzing the semantic analysis of the hot spot content and combining the comments related to the current hot spot. The industry relevance matching unit identifies the industry to which the current hot spot information belongs based on the semantic analysis results of the hot spot information. The heat value is a comprehensive evaluation based on the click volume, attention, and evaluation volume of the current hot spot information.

[0066] In this embodiment, when performing the level division, the sentiment tendency, the industry relevance and the heat degree are taken as three basic factors, and a corresponding coefficient is given to each factor, and the formula is as follows: L = aQ + bH + cR; Wherein, L represents the sum value, a, b and c are the corresponding coefficients given, Q represents the sentiment value, H represents the industry relevance value, and R represents the heat value; After the value of each hot information is given based on the above three dimensions and the corresponding coefficients, the sum is performed; In this embodiment, different levels are given based on the numerical range of different sum results, such as eight levels are given based on different numerical ranges, and the top four or five levels of hot spots are selected as the fusion objects.

[0067] As a preferred embodiment: the delivery content processing module includes a content reorganization unit and a hot spot visual element adding unit, the content reorganization unit is used to insert the extracted core word into the enterprise's own media content, and the hot spot visual element adding unit is used to add visual elements related to the current hot spot to the enterprise's own media content, and generate new media delivery content.

[0068] As another preferred embodiment: the hot spot information processing module includes a hot spot information slicing unit, the hot spot information slicing unit is used to slice the received hot spot information and identify and extract key hot spot segments, and the delivery content processing module is used to implant the enterprise's own media content into the key hot spot segments to generate new media delivery content, or the delivery content processing module extracts segments or special effects from the key hot spot segments and applies them to the enterprise's own media content to generate new media delivery content.

[0069] The output of this embodiment is as follows: The original video title of the customer is: "Recommendation of new products in local barbecue shop" The title after fusion is: "Zibo barbecue is on fire! Teach you how to reproduce the same carbon barbecue secret recipe @ XX barbecue shop".

[0070] Hot event: a game character dance is very popular Customer industry: children's dance training Generated content: students jumping the dance segment, agency logo watermark and subtitle "Open reservation for popular dance offline class".

[0071] In this embodiment, after the new fused media content is delivered, it is also monitored by a test module, the test module is used to automatically monitor the original delivered media content and the new delivered media content fused with hot spots, and compare and analyze the click volume and lead information conversion rate of the two.

[0072] In this embodiment, the hotspot information monitoring processing system further comprises a dynamic delivery strategy module, the dynamic delivery strategy module is used for automatically associating the hotspot label with the customer industry attribute, generating a delivery matrix, the delivery matrix comprises a hotspot dimension, a customer dimension and a delivery action, when the hotness value of the applied hotspot information exceeds the set hotness value, and the industry matching degree exceeds the correlation degree set value, the delivery action is to generate new media delivery content, and the new media delivery content is delivered on the whole Internet platform; when the click volume rising rate of the new media content delivered in a period of time exceeds the set rate value, and the historical customer click volume is greater than the click volume set value, the delivery action is to add delivery budget; when it is monitored in a period of time that the participating customer group is a specific age group, and the age group coincides with the target customer group, the Internet platform matched with the corresponding specific age group is delivered as the focus.

[0073] The dynamic delivery matrix is as shown in Table 1. Table 1: Dynamic delivery matrix The delivery content processing module further comprises a lead information conversion module, the lead information conversion module is used for automatically performing page jumping when the customer group clicks, comments and focuses on the new media content delivered, and instantly obtaining the gifts or services delivered by the enterprise, and the lead information conversion module automatically captures the lead information of the customer group in the process that the customer group obtains the gifts or services of the enterprise. After the customer group clicks, comments and focuses on the new media content delivered, the page is automatically jumped, the gifts or services delivered by the enterprise are instantly obtained, and in this process, the lead information conversion module is used to capture the lead information of the customer group in time, so that the purpose of increasing the lead information obtained by the enterprise is achieved.

[0074] The hotspot information monitoring system of this embodiment further comprises a preset template library: including a title template, a cover template and a tail conversion component.

[0075] The “tail conversion component” in the prefabricated template library is a kind of pre-designed dynamic or static template element, which is used for being inserted at the end of video / text content, guiding users to perform specific actions (such as clicking links, filling in forms, focusing on accounts, etc.). These components contain buttons, two-dimensional codes, text prompts and other visual elements, and are connected with the lead collection system of the enterprise (such as CRM).

[0076] Function: guide conversion: attract users to click through eye-catching design (such as “limited time to get” “immediately consult”); automatically capture leads: automatically trigger the lead information collection process (such as pop-up forms, authorized user information) after the user clicks; template diversification: provide multiple styles of templates (such as technology, warm style) according to industry and hotspot type, support custom text and visual elements; data statistics: record the exposure, click volume and conversion rate of the component, and optimize the delivery strategy.

[0077] The embodiment also discloses a hotspot information monitoring processing method, and specifically comprises the following contents. Monitoring hotspot information of the current Internet; Collecting the monitored hotspot information; Analyzing the emotional tendency of the currently collected hotspot information, screening out hotspot information with positive emotions and emotional values, screening out hotspot information related to the industry and calculating the matching degree, calculating the heat value of the hotspot information in a historical time period, and classifying the hotspot information based on the heat value, industry matching degree and emotional value of the hotspot information; Performing semantic analysis and core word extraction on hotspot information with a grade higher than a preset grade; inserting the extracted core word into the enterprise's own media content, adding visual elements related to the current hotspot to the enterprise's own media content, and generating new media delivery content; Or slicing the received hotspot information, identifying and extracting key hotspot segments, implanting the enterprise's own media content into the key hotspot segments to generate new media delivery content, or extracting segments or special effects from the key hotspot segments and applying them to the enterprise's own media content to generate new media delivery content.

[0078] The embodiment also discloses an application of a hotspot information monitoring processing system, which is applied to a scenario in which an enterprise fuses the obtained hotspot information with its own media content, delivers the fused media content, and thereby obtains the clue information of the customer group of the enterprise.

[0079] The preferred embodiments of the present application are described in detail in combination with the drawings, and typical known structures and known common knowledge technologies are not described in detail herein. Those skilled in the art can improve and implement the technical solutions of the present application based on their own abilities under the guidance of the present embodiments, and some typical known structures, known methods or known common knowledge technologies should not be an obstacle for those skilled in the art to implement the present application.

[0080] The scope of protection of the present application should be subject to the content of its claims, and the content recorded in the summary, detailed description and drawings of the specification is used to explain the claims.

[0081] Within the technical concept of the present application, several modifications can be made to the specific embodiments of the present application, and the specific embodiments after the modifications should also be considered within the protection scope of the present application.

Claims

1. A hotspot information monitoring processing system, characterized by: The application comprises a hotspot information monitoring module, a hotspot information collecting module, a hotspot information processing module, and a content delivery processing module. The hotspot information monitoring module is used to monitor the current internet hotspot information. The hotspot information collecting module is used to collect the monitored hotspot information. The hotspot information processing module is used to process the collected hotspot information and generate a processing result. The content delivery processing module combines the processing result to process the enterprise's own media content and generate new media delivery content. The application further comprises a dynamic delivery strategy module, which is used to automatically associate the hotspot label with the customer industry attribute, generate a delivery matrix, and include the hotspot dimension, customer dimension, and delivery action in the delivery matrix. When the hotness value of the applied hotspot information exceeds the set hotness value and the industry matching degree exceeds the set correlation value, the delivery action is to generate new media delivery content and deliver it on the internet platform. When the click rate of the delivered new media content in a period of time exceeds the set rate value and the historical customer click value is greater than the set click value, the delivery action is to add the delivery budget. When the participating customer group is a specific age group in a period of time and the age group coincides with the target customer group, the application focuses on delivering the internet platform that matches the specific age group.

2. The hotspot information monitoring processing system according to claim 1, characterized by: The hotspot information processing module further comprises an emotional tendency analysis unit, an industry relevance matching unit, and a hotness value calculation unit. The emotional tendency analysis unit is used to analyze the emotional tendency of the collected hotspot information and filter out the hotspot information with positive emotions and emotional values. The industry relevance matching unit is used to filter out the hotspot information related to the industry and calculate the matching degree. The hotspot information hotness value calculation unit is used to calculate the hotness value of the hotspot information in the historical period. The hotspot information processing module classifies the hotspot information based on the hotness value, industry matching degree, and emotional value of the hotspot information.

3. The hotspot information monitoring processing system according to claim 1, characterized by: The hotspot information processing module further comprises a hotspot semantic analysis unit and a core word extraction unit. The hotspot semantic analysis unit is used to perform semantic analysis on the filtered hotspot information. The core word extraction unit is used to extract the core words after semantic analysis.

4. The hotspot information monitoring processing system according to claim 3, characterized by: The content delivery processing module comprises a content reorganization unit and a hotspot visual element adding unit. The content reorganization unit is used to insert the extracted core words into the enterprise's own media content. The hotspot visual element adding unit is used to add visual elements related to the current hotspot to the enterprise's own media content and generate new media delivery content.

5. The hotspot information monitoring processing system of claim 1, wherein: The hotspot information processing module comprises a hotspot information slicing unit, which is used to slice the received hotspot information and identify and extract key hotspot segments. The content delivery processing module is used to implant the enterprise's own media content into the key hotspot segments to generate new media delivery content. Alternatively, the content delivery processing module extracts segments or special effects from the key hotspot segments and applies them to the enterprise's own media content to generate new media delivery content.

6. The hotspot information monitoring processing system according to claim 2, characterized by The analysis method of the industry relevance matching unit is as follows: Construct an industry keyword library; Calculate the cosine similarity between the hot text and the keyword library; Matching degree calculation formula: In the formula, ki is the i-th industry keyword, T is the hot text, TF-IDF(ki) is the TF-IDF weight of the keyword ki, and Sim(ki, T) is the word vector similarity between the keyword ki and the hot text T.

7. The hotspot information monitoring processing system according to claim 4 or 5, characterized by: It also includes a test module for automatically monitoring the original media content and the new media content combined with the hot spot, and comparing and analyzing the click volume and lead information conversion rate of the two.

8. The hotspot information monitoring processing system according to claim 6, characterized by: The heat value calculation unit calculates the heat value in the historical time period based on the time decay model, and integrates the click volume, comment volume and forwarding volume indicators. The formula is as follows: Where: R represents the heat value, C represents the click volume, Rf represents the forwarding volume, and M represents the comment volume; Wc, Wr, and Wm are the weights of the click volume, forwarding volume, and comment volume, respectively, reflecting the contribution of each indicator to the heat; λ is the decay factor, controlling the rate of heat decay over time; t is the time interval, starting from the time of publishing the hot information.

9. The hotspot information monitoring processing system according to claim 6, characterized by: The heat value calculation unit calculates the heat value in the historical time period based on the time decay model, and integrates the click volume and forwarding volume indicators. The formula is as follows: Wherein, H t is the hot value, C i is the click volume, R i is the forwarding volume, w c , w r is the weight, and λ is the attenuation factor.

10. A method of hotspot information monitoring processing, characterized by, Specifically, the following content is included: Monitoring the current Internet hot information; Collecting the monitored hot information; Analyzing the emotional tendency of the current collected hot information, and screening out hot information with positive emotions and emotional values, screening out hot information related to the industry and calculating the matching degree, calculating the heat value of the hot information in the historical time period, and classifying the hot information based on the heat value, industry matching degree and emotional value of the hot information; For hot information with a grade higher than a preset grade, perform semantic analysis and core word extraction; insert the extracted core words into the enterprise's own media content, and add visual elements related to the current hot spot to the enterprise's own media content, and generate new media content; Or slice the received hot information and identify and extract key hot spots, and implant the enterprise's own media content into the key hot spot segments to generate new media content, or extract segments or special effects from the key hot spot segments and apply them to the enterprise's own media content to generate new media content; The dynamic delivery strategy module automatically associates hot spot tags with customer industry attributes, generates a delivery matrix, and the delivery matrix includes hot spot dimensions, customer dimensions and delivery actions. When the heat value of the applied hot information exceeds the set heat value, and the industry matching degree exceeds the set value of the correlation degree, the delivery action is to generate new media content and deliver it on the Internet platform. When the click volume of the new media content delivered in a period of time exceeds the set rate value, and the historical customer click volume is greater than the set value of the click volume, the delivery action is to add delivery budget. When a specific age group of customers is monitored in a period of time, and the age group of customers coincides with the target customer group, the Internet platform matching the specific age group of customers is focused on.

Images