Short video popularity prediction method and device, equipment and storage medium

Abstract

The invention discloses a short video popularity prediction method and device, equipment and a storage medium, and relates to the technical field of computers. The method comprises the following steps: acquiring data information associated with a short video to be predicted from a plurality of data sources; determining interaction feature information and text feature information of the to-be-predicted short video based on the data information; and predicting based on the interaction feature information and the text feature information to obtain a popularity value of the to-be-predicted short video. Multi-aspect data related to the short video can be widely collected, analysis one-sidedness caused by a single data source is avoided, and then popularity related information of the short video is deeply mined from two dimensions of interactive feature information and text feature information. Therefore, the actual popularity degree and the potential development trend of the short video can be mastered more accurately by the popularity prediction, and the comprehensiveness and the accuracy of the popularity prediction of the short video are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computers, and particularly relates to a short video heat prediction method and device, equipment and a storage medium. BACKGROUND

[0002] In the rapidly developing short video entertainment market today, short videos have rapidly risen and occupied an important market share due to their shortness, compactness and fast rhythm. A short video with high heat can not only attract a large number of audiences and improve the user activity and stickiness of a platform, but also bring considerable advertising revenue. Therefore, accurately predicting the heat of a short video is of great significance for a platform to formulate content operation strategies and reasonably allocate promotion resources.

[0003] At present, most existing heat prediction methods rely on the analysis of data in a single dimension. This method is too one-sided and ignores the data generated by the interactive behaviors of audiences in the watching process, and thus cannot accurately grasp the actual popularity and potential development trend of a short video, making it difficult to make a comprehensive heat prediction and resulting in low accuracy of the prediction result. SUMMARY

[0004] The present application provides a short video heat prediction method, device, equipment and storage medium, which solves the problem that most existing heat prediction methods rely on the analysis of data in a single dimension and are difficult to make a comprehensive heat prediction, resulting in a large deviation between the prediction result and the actual situation, and improves the comprehensiveness and accuracy of the prediction.

[0005] In a first aspect, the present application provides a short video heat prediction method, comprising: obtaining data information associated with a to-be-predicted short video from multiple data sources; determining interactive feature information and text feature information of the to-be-predicted short video based on the data information; predicting a heat value of the to-be-predicted short video based on the interactive feature information and the text feature information.

[0006] Optionally, the step of obtaining data information associated with a to-be-predicted short video from multiple data sources comprises: determining a keyword related to the to-be-predicted short video, and expanding the keyword to obtain a keyword set; determining an interface corresponding to each data source; obtaining data information associated with the to-be-predicted short video from each data source through the interface corresponding to each data source based on the keyword set.

[0007] Optionally, the step of obtaining data information associated with the short video to be predicted from each of the data sources based on the keyword set through the interface corresponding to each of the data sources includes: Based on the keyword set and the interface corresponding to the data source, construct the request parameters corresponding to each data source; Based on the interface protocol of each data source and the request parameters, construct a query request to access the interface of each data source; The query request is sent to the corresponding interface of the data source, and the returned response data is received to obtain the data information associated with the short video to be predicted.

[0008] Optionally, determining the interaction feature information and text feature information of the short video to be predicted based on the data information includes: The interactive data information corresponding to the short video to be predicted is obtained from the data information, and the interactive feature information of the short video to be predicted is determined based on the interactive data information, wherein the interactive data information includes any one or more of the following: number of likes, number of comments, number of plays, number of shares, and number of searches. The text data information corresponding to the short video to be predicted is obtained from the data information, wherein the text data information includes comment text information and bullet screen text information; Based on the comment text information and the bullet screen text information, the text feature information of the video to be predicted is determined.

[0009] Optionally, predicting the popularity value of the short video to be predicted based on the interaction feature information and text feature information includes: The interaction feature information is quantized to obtain the interaction quantization value; Sentiment analysis is performed on the text feature information, and a sentiment quantification value is determined based on the analysis results; Based on the interaction quantification value and the emotion quantification value, the popularity value of the short video to be predicted is obtained.

[0010] Optionally, performing sentiment analysis on the text feature information and determining a sentiment quantification value based on the analysis results includes: Sentiment analysis is performed on the text feature information to determine the sentiment tendency and sentiment intensity of the text feature information; The emotional quantification value is determined based on the emotional tendency and emotional intensity.

[0011] Optionally, the step of quantizing the interaction feature information to obtain the interaction quantization value includes: Each interaction dimension index in the interaction feature information is normalized to obtain the corresponding normalized value; The normalized values ​​corresponding to each of the interaction dimension indicators are weighted and summed to obtain the interaction quantification value.

[0012] Secondly, this application provides a short video popularity prediction device, comprising: The information acquisition module is used to acquire data information related to the short video to be predicted from multiple data sources; The feature information determination module is used to determine the interactive feature information and text feature information of the short video to be predicted based on the data information; The popularity prediction module is used to predict the popularity value of the short video to be predicted based on the interaction feature information and text feature information.

[0013] Thirdly, this application provides a short video popularity prediction device, including: One or more processors; A memory that stores one or more programs that, when executed by one or more processors, enable the one or more processors to implement the short video popularity prediction method as described in the first aspect.

[0014] Fourthly, this application provides a storage medium containing computer-executable instructions, which, when executed by a computer processor, are used to perform the short video popularity prediction method as described in the first aspect.

[0015] In summary, compared with the prior art, the effective effects of the technical solution provided in this application include at least the following: In this application, data information associated with the short video to be predicted is obtained from multiple data sources; based on the data information, the interactive feature information and textual feature information of the short video to be predicted are determined; and based on the interactive feature information and textual feature information, the popularity value of the short video to be predicted is predicted. In the above technical means, by obtaining data information associated with the short video to be predicted from multiple data sources, a wide range of data related to short videos can be gathered, avoiding the one-sidedness of analysis caused by a single data source. Furthermore, by deeply mining the popularity-related information of short videos from the two dimensions of interactive feature information and textual feature information, the popularity prediction can more accurately grasp the actual popularity and potential development trend of short videos, improving the comprehensiveness and accuracy of short video popularity prediction. Attached Figure Description

[0016] Figure 1 This is a flowchart of a short video popularity prediction method provided in an embodiment of this application; Figure 2 This is a flowchart illustrating the process of obtaining data information associated with the short video to be predicted, as provided in an embodiment of this application. Figure 3This is a flowchart illustrating the process of obtaining data information associated with the short video to be predicted based on a keyword set, as provided in an embodiment of this application. Figure 4 This is a flowchart of determining the interactive feature information and text feature information of the short video to be predicted, provided in an embodiment of this application; Figure 5 This is a flowchart illustrating the process of predicting the popularity value of a short video to be predicted, provided in an embodiment of this application. Figure 6 This is a flowchart provided in an embodiment of the present application for quantifying interactive feature information to obtain interactive quantification values; Figure 7 This is a flowchart of determining the quantification value of emotion provided in an embodiment of this application; Figure 8 This is a schematic diagram of the structure of a short video popularity prediction device provided in an embodiment of this application; Figure 9 This is a schematic diagram of the structure of a short video popularity prediction device provided in an embodiment of this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this application clearer, specific embodiments of this application will be described in further detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely for explaining this application and not for limiting it. It should also be noted that, for ease of description, only the parts relevant to this application are shown in the drawings, not all of them. Before discussing exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe operations (or steps) as sequential processes, many of these operations can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. A process can be terminated when its operation is completed, but it may also have additional steps not included in the drawings. A process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0018] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0019] To address the technical problem that existing methods for predicting short video popularity largely rely on single-dimensional data analysis, which is too one-sided and ignores data generated by audience interaction during viewing, thus failing to accurately grasp the actual popularity and potential development trends of short videos and making comprehensive popularity prediction difficult, resulting in low accuracy, this application provides a method for predicting short video popularity. By acquiring data information related to the short video to be predicted from multiple data sources, it can widely gather multifaceted data related to short videos, avoiding the one-sidedness of analysis caused by a single data source. Furthermore, it delves into the popularity-related information of short videos from two dimensions: interactive feature information and text feature information, enabling popularity prediction to more accurately grasp the actual popularity and potential development trends of short videos, thereby improving the comprehensiveness and accuracy of short video popularity prediction.

[0020] The short video popularity prediction method provided in this embodiment can be executed by a short video popularity prediction device, which can be implemented by software and / or hardware. The short video popularity prediction device can be composed of two or more physical entities, or it can be composed of a single physical entity.

[0021] The short video popularity prediction device is equipped with at least one type of operating system, including but not limited to Android, Linux, and Windows. The device can install at least one application based on the operating system; this application can be a built-in application of the operating system or an application downloaded from a third-party device or server. In this embodiment, the short video popularity prediction device has at least one application capable of executing the short video popularity prediction method.

[0022] The technical solutions provided by the embodiments of this application are described in detail below with reference to the accompanying drawings.

[0023] Figure 1 A flowchart of a short video popularity prediction method provided in an embodiment of this application is given. Please refer to it. Figure 1 The short video popularity prediction method specifically includes: S110. Obtain data information associated with the short video to be predicted from multiple data sources.

[0024] The data source refers to various information carriers that can provide data related to the short video to be predicted, such as physical databases and online information platforms. Optional data sources include: the short video platform's own system (referring to the publishing platform where the short video to be predicted is located and other similar short video distribution platforms), social media platforms (network platforms with social attributes), search engine services (platforms that provide information retrieval services, etc.), news and information platforms, and short video platform ranking systems, etc.

[0025] The data information associated with the short video to be predicted refers to a collection of various data that can directly or indirectly reflect the content characteristics, user feedback, dissemination scope and impact of the short video; optionally, the data information associated with the short video to be predicted includes user interaction data for the short video (including the number of likes, comments, shares, favorites, downloads, viewing completion rate and number of bullet comments sent), publication time data, specific comment texts and bullet comment texts of users for the short video, etc.

[0026] This application obtains data information related to the short video to be predicted from multiple data sources. By integrating data from different channels, it provides a richer and more complete data foundation for subsequent popularity prediction, enabling the prediction results to more comprehensively reflect the various factors affecting the popularity of short videos and reducing prediction bias caused by incomplete data.

[0027] For example, Figure 2 This is a flowchart illustrating the acquisition of data information associated with the short video to be predicted, provided in an embodiment of this application; please refer to [link / reference]. Figure 2 The steps for obtaining data information associated with the short video to be predicted include S1101-S1103: S1101. Determine keywords related to the short video to be predicted, and expand the keywords to obtain a keyword set.

[0028] Keywords refer to words or phrases that describe the core content, theme, characteristics, specific characters, and key elements of the short video to be predicted. These keywords serve as crucial evidence for searching, identifying, and extracting data related to the short video, helping to accurately obtain data information associated with the short video from various data sources.

[0029] Furthermore, after identifying the keywords, they can be expanded. Expansion methods can include synonym replacement, near-synonym expansion, and related word association to comprehensively cover various forms of expression related to short videos and improve the comprehensiveness of data acquisition.

[0030] S1102. Determine the interface corresponding to each of the data sources.

[0031] In this context, an interface refers to the access point provided by the data source for data interaction. It serves as the channel for data transmission between the data source and external systems. Optionally, the interface can be an API (Application Programming Interface), which typically consists of a set of predefined functions, classes, data structures, and communication protocols. By calling these API functions and sending requests to the data source according to specified parameters and methods, the required data can be obtained. An interface can also be a web crawler interface. For data sources that do not provide API interfaces or data export interfaces, web crawling technology can be used to obtain data. A web crawler interface refers to the entry address of the webpage or network service of the data source. By simulating the user's browser behavior, requests are sent to the web crawler interface to obtain webpage content or other forms of data.

[0032] S1103. Based on the keyword set, obtain data information associated with the short video to be predicted from each of the data sources through the interface corresponding to each data source.

[0033] Optional, Figure 3 This is a flowchart illustrating the process of obtaining data information associated with the short video to be predicted based on a keyword set, as provided in an embodiment of this application. For example... Figure 3 As shown, the steps for obtaining data information associated with the short video to be predicted based on the keyword set specifically include S11031-S11033: S11031. Based on the keyword set and the interface corresponding to the data source, construct the request parameters corresponding to each data source.

[0034] Request parameters refer to information used to pass specific instructions to the interface, guiding the data source to filter and return data that meets the requirements according to the specified rules and requirements.

[0035] Specifically, determine the specific requirements of each data source interface for request parameters, including parameter names, data types, etc., and then fill in the words from the expanded keyword set into the corresponding request parameter fields as required.

[0036] S11032. Construct a query request to access the interface of each data source according to the interface protocol of each data source and the request parameters.

[0037] Among them, the interface protocol is the communication standard that the data source must follow when interacting with external systems.

[0038] S11033. Send the query request to the interface of the corresponding data source, receive the returned response data, and obtain the data information associated with the short video to be predicted.

[0039] In this embodiment, since the raw data obtained from different data sources often varies in quality, it is necessary to clean the acquired data. Optionally, data cleaning may include correcting erroneous data and supplementing missing data. Furthermore, since data obtained from multiple data sources may differ in format and structure—for example, data from some data source platforms may be returned in JSON format, while data from others may be provided in XML format; field names may also differ between data sources—it is also necessary to perform format conversion on the data information. Through format conversion, the data information from different data sources is converted into a unified standard format to facilitate subsequent data processing and analysis.

[0040] S120. Based on the data information, determine the interactive feature information and text feature information of the short video to be predicted.

[0041] Interactive feature information refers to a set of quantitative features that reflect user engagement, dissemination scope, and popularity. Textual feature information refers to information extracted from user comments and bullet screen texts on short videos that reflects user sentiment, focus, and discussion topics. This information helps to gain a deeper understanding of users' true opinions and feelings about short videos.

[0042] Optional, Figure 4 This is a flowchart illustrating the determination of interactive and textual feature information of a short video to be predicted, provided in an embodiment of this application. For example... Figure 4 As shown, the steps for determining the interactive feature information and textual feature information of the short video to be predicted specifically include S1201-S1203: S1201. Obtain the interaction data information corresponding to the short video to be predicted from the data information, and determine the interaction feature information of the short video to be predicted based on the interaction data information, wherein the interaction data information includes any one or more of the following: number of likes, number of comments, number of plays, number of shares, and number of searches.

[0043] Specifically, the data of each interaction dimension indicator obtained from different data sources are integrated, statistically analyzed, and the growth rate and change pattern of each interaction dimension indicator are determined to obtain interaction feature information.

[0044] S1202. Obtain the text data information corresponding to the short video to be predicted from the data information, wherein the text data information includes comment text information and bullet screen text information.

[0045] Among them, bullet screen text information refers to the scrolling bullet screen text information on the short video screen; comment text information refers to the information in the comment section of the short video.

[0046] S1203. Based on the comment text information and the bullet screen text information, determine the text feature information of the video to be predicted.

[0047] S130. Based on the interactive feature information and text feature information, predict the popularity value of the short video to be predicted.

[0048] The popularity score is a comprehensive quantitative indicator that measures the level of attention and popularity a short video receives from users. It takes into account user interaction behaviors (such as likes, comments, and shares) and user emotional feedback on the short video content (such as positive, negative, or neutral emotions), providing a direct reflection of the short video's popularity and influence. A higher popularity score indicates a more popular short video, a wider reach, and a stronger appeal to users.

[0049] Optional, Figure 5 This is a flowchart illustrating the process of predicting the popularity value of a short video to be predicted, provided in an embodiment of this application. For example... Figure 5 As shown, the specific steps for obtaining the popularity value of the short video to be predicted include S1301-S1303: S1301. The interactive feature information is quantized to obtain the interactive quantization value.

[0050] The interaction quantification score is a quantitative measure of users' overall performance in interacting with short videos. It integrates and quantifies multiple interaction metrics (such as likes, comments, and shares) to evaluate the popularity and user engagement of short videos. A higher interaction quantification score indicates that the video is more likely to stimulate user enthusiasm and thus has higher popularity.

[0051] Optional, Figure 6 This is a flowchart illustrating the process of quantifying interactive feature information to obtain interactive quantification values, as provided in an embodiment of this application. For example... Figure 6 As shown, this is as follows Figure 6 As shown, the step of quantifying the interaction feature information to obtain the interaction quantization value specifically includes S13011-S13012: S13011. Normalize each interaction dimension index in the interaction feature information to obtain the corresponding normalized value.

[0052] For each interaction dimension indicator, its normalization range is determined. Setting the normalization range to [0, 1] can map data of different dimensions to the same scale, which is convenient for subsequent weighted summation calculation. Optionally, this application can use a linear normalization method to map the value of each interaction dimension indicator to the normalization range.

[0053] S13012. The normalized values ​​corresponding to each of the interaction dimension indicators are weighted and summed to obtain the interaction quantification value.

[0054] S1302. Perform sentiment analysis on the text feature information and determine the sentiment quantification value based on the analysis results.

[0055] Among them, the sentiment quantification value is a quantitative value that reflects the user's emotional attitude towards short videos, obtained by performing sentiment analysis on text feature information (comment text, bullet screen text).

[0056] Optional, Figure 7 This is a flowchart illustrating the determination of emotion quantification values ​​provided in an embodiment of this application. For example... Figure 7 As shown, this is as follows Figure 7 As shown, the steps for determining the emotional quantification value specifically include S13021-S13022: S13021. Perform sentiment analysis on the text feature information to determine the sentiment tendency and sentiment intensity of the text feature information; Sentiment orientation refers to the user's emotional attitude towards short videos, which can be divided into three types: positive, negative, and neutral. A positive sentiment orientation indicates that the user holds positive attitudes such as liking, satisfaction, and support for the short videos; a negative sentiment orientation indicates that the user has negative emotions such as aversion, dissatisfaction, and criticism; a neutral sentiment orientation means that the user's emotional attitude is relatively neutral, with neither obvious liking nor obvious aversion. Sentiment intensity refers to the degree of the user's sentiment orientation, that is, the intensity of the emotion. For example, a high positive sentiment intensity indicates that the user likes the short videos very much and is extremely satisfied; a low positive sentiment intensity indicates that the user likes the short videos to some extent, but not particularly strongly.

[0057] Furthermore, this application can determine the sentiment tendency and intensity of text features by inputting text feature information into a trained sentiment model. For example, a machine learning-based method can be used to train the sentiment classification model, such as using Naive Bayes, Support Vector Machine (SVM), Recurrent Neural Network (RNN), etc. During training, a certain amount of sentiment-annotated text data is prepared as training data, with annotations including sentiment tendency and intensity; for example, the sentiment tendency of the text "This short video is so interesting! I love it!" is labeled as positive, and the intensity is the strongest level. Then, the text data is converted into a feature vector form acceptable to the model, such as using word embedding technology to map words to a high-dimensional vector space to obtain the semantic feature representation of the text; the selected sentiment model is trained using the annotated training data, and the model parameters are adjusted through optimization algorithms (such as gradient descent) so that the model can accurately predict the sentiment tendency and intensity of the text.

[0058] S13022. Determine the emotional quantification value based on the emotional tendency and emotional intensity.

[0059] Determine the sentiment weight corresponding to each sentiment tendency (positive, negative, and neutral), and determine the sentiment intensity value corresponding to different intensity levels. For each comment text and bullet screen text, multiply the determined sentiment weight corresponding to the sentiment tendency by the sentiment intensity value corresponding to the sentiment intensity to obtain the initial sentiment quantification value of the text. Then, sum the initial sentiment quantification values ​​of all texts and divide them by the total number of texts involved in the calculation to obtain the average sentiment quantification value as the sentiment quantification value.

[0060] S1303. Based on the interaction quantification value and the emotion quantification value, the popularity value of the short video to be predicted is obtained.

[0061] The weights of interaction quantification and sentiment quantification in popularity prediction are determined, and the weighted sum of these two values ​​is used to predict the short video's popularity. This popularity prediction based on the aforementioned interaction and textual features avoids the problem of significant deviations between predictions and reality caused by single-dimensional analysis, improving the comprehensiveness and accuracy of the prediction results. Furthermore, it provides a more reliable basis for platforms to formulate content operation strategies and rationally allocate promotional resources, which is conducive to improving the platform's overall benefits, such as user activity, stickiness, and advertising revenue.

[0062] In summary, the short video popularity prediction method provided in this application involves acquiring data information associated with the short video to be predicted from multiple data sources; determining the interactive and textual feature information of the short video to be predicted based on the data information; and predicting the popularity value of the short video to be predicted based on the interactive and textual feature information. In the above technical means, by acquiring data information associated with the short video to be predicted from multiple data sources, a wide range of data related to short videos can be gathered, avoiding the one-sidedness of analysis caused by a single data source. Furthermore, by deeply mining the popularity-related information of short videos from the two dimensions of interactive and textual feature information, the popularity prediction can more accurately grasp the actual popularity and potential development trend of short videos, improving the comprehensiveness and accuracy of short video popularity prediction.

[0063] Figure 8 This is a schematic diagram of a short video popularity prediction device provided in an embodiment of this application. (Reference) Figure 8 The short video popularity prediction device provided in this embodiment specifically includes: an information acquisition module 21, a feature information determination module 22, and a popularity prediction module 23.

[0064] Information acquisition module 21 is used to acquire data information related to the short video to be predicted from multiple data sources; The feature information determination module 22 is used to determine the interactive feature information and text feature information of the short video to be predicted based on the data information; The popularity prediction module 23 is used to predict the popularity value of the short video to be predicted based on the interactive feature information and text feature information.

[0065] Based on the above embodiments, the information acquisition module 21 includes: The keyword set determination unit is used to determine keywords related to the short video to be predicted, and to expand the keywords to obtain a keyword set; An interface determination unit is used to determine the interface corresponding to each of the data sources; The information acquisition unit is used to acquire data information associated with the short video to be predicted from each of the data sources through the interface corresponding to each of the data sources, based on the keyword set.

[0066] Based on the above embodiments, the information acquisition unit includes: The parameter construction subunit is used to construct request parameters corresponding to each of the data sources based on the keyword set and the interface corresponding to the data source; A request construction subunit is used to construct a query request to access the interface of each of the data sources based on the interface protocol of each data source and the request parameters. The information acquisition subunit is used to send the query request to the interface of the corresponding data source and receive the returned response data to obtain data information associated with the short video to be predicted.

[0067] Based on the above embodiments, the feature information determination module 22 includes: An interactive feature information determination unit is used to obtain interactive data information corresponding to the short video to be predicted from the data information, and determine the interactive feature information of the short video to be predicted based on the interactive data information, wherein the interactive data information includes any one or more of likes, comments, plays, shares and searches; A text information acquisition unit is used to acquire text data information corresponding to the short video to be predicted from the data information, wherein the text data information includes comment text information and bullet screen text information; The text feature information determination unit is used to determine the text feature information of the video to be predicted based on the comment text information and the bullet screen text information.

[0068] Based on the above embodiments, the heat prediction module 23 includes: An interaction quantization value determination unit is used to quantize the interaction feature information to obtain an interaction quantization value; The sentiment quantification value determination unit is used to perform sentiment analysis on the text feature information and determine the sentiment quantification value based on the analysis results; The popularity prediction unit is used to predict the popularity value of the short video to be predicted based on the interaction quantification value and the emotion quantification value.

[0069] Based on the above embodiments, the emotion quantification value determination unit includes: The sentiment determination subunit is used to perform sentiment analysis on the text feature information to determine the sentiment tendency and sentiment intensity of the text feature information. The emotion quantification value determination subunit is used to determine the emotion quantification value based on the emotion tendency and emotion intensity.

[0070] Based on the above embodiments, the interactive quantization value determination unit includes: The normalization processing subunit is used to normalize each interaction dimension index in the interaction feature information to obtain the corresponding normalized value. The interaction quantification value determination subunit is used to perform a weighted summation of the normalized values ​​corresponding to each interaction dimension indicator to obtain the interaction quantification value.

[0071] In summary, the short video popularity prediction device provided in this application obtains data information associated with the short video to be predicted from multiple data sources; determines the interactive feature information and text feature information of the short video to be predicted based on the data information; and predicts the popularity value of the short video to be predicted based on the interactive feature information and text feature information. In the above technical means, by obtaining data information associated with the short video to be predicted from multiple data sources, a wide range of data related to short videos can be collected, avoiding the one-sidedness of analysis caused by a single data source. Furthermore, by deeply mining the popularity-related information of short videos from the two dimensions of interactive feature information and text feature information, the popularity prediction can more accurately grasp the actual popularity and potential development trend of short videos, improving the comprehensiveness and accuracy of short video popularity prediction.

[0072] The short video popularity prediction device provided in this application embodiment can be used to execute the short video popularity prediction method provided in the above embodiment, and has corresponding functions and beneficial effects.

[0073] Figure 9 This is a schematic diagram of the structure of a short video popularity prediction device provided in an embodiment of this application, with reference to... Figure 9The short video popularity prediction device includes a processor 31, a memory 32, a communication device 33, an input device 34, and an output device 35. The number of processors 31 and the number of memories 32 in the short video popularity prediction device can be one or more. The processor 31, memory 32, communication device 33, input device 34, and output device 35 of the short video popularity prediction device can be connected via a bus or other means.

[0074] The memory 32, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as program instructions / modules corresponding to the short video popularity prediction method in any embodiment of this application (e.g., the information acquisition module 21, feature information determination module 22, and popularity prediction module 23 in the short video popularity prediction device). The memory 32 may primarily include a program storage area and a data storage area. The program storage area may store the operating system and at least one application program required for a function; the data storage area may store data created based on the use of the device, etc. Furthermore, the memory 32 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the device via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0075] The communication device 33 is used for data transmission.

[0076] The processor 31 executes various functional applications and data processing of the device by running software programs, instructions and modules stored in the memory 32, thereby realizing the above-mentioned short video popularity prediction method.

[0077] Input device 34 can be used to receive input digital or character information, and to generate key signal inputs related to user settings and function control of the device. Output device 35 may include display devices such as a display screen.

[0078] The short video popularity prediction device provided above can be used to execute the short video popularity prediction method provided in the above embodiments, and has corresponding functions and beneficial effects.

[0079] This application embodiment also provides a storage medium containing computer-executable instructions. When executed by a computer processor, the computer-executable instructions are used to perform a short video popularity prediction method. The short video popularity prediction method includes: obtaining data information associated with the short video to be predicted from multiple data sources; determining the interactive feature information and text feature information of the short video to be predicted based on the data information; and predicting the popularity value of the short video to be predicted based on the interactive feature information and text feature information.

[0080] Storage medium – any type of memory device or storage device. The term “storage medium” is intended to include: mounting media, such as CD-ROM, floppy disk, or magnetic tape devices; computer system memory or random access memory, such as DRAM, DDR RAM, SRAM, EDO RAM, Rambus RAM, etc.; non-volatile memory, such as flash memory, magnetic media (e.g., hard disk or optical storage); registers or other similar types of memory elements, etc. Storage medium may also include other types of memory or combinations thereof. Furthermore, storage medium may reside in a first computer system in which the program is executed, or it may reside in a different second computer system connected to the first computer system via a network (such as the Internet). The second computer system can provide program instructions to the first computer for execution. The term “storage medium” can include two or more storage media residing in different locations (e.g., in different computer systems connected via a network). Storage medium may store program instructions (e.g., specifically implemented as a computer program) executable by one or more processors.

[0081] Of course, the computer-executable instructions provided in the embodiments of this application are not limited to the short video popularity prediction method described above, but can also execute related operations in the short video popularity prediction method provided in any embodiment of this application.

[0082] The short video popularity prediction device, storage medium, and short video popularity prediction equipment provided in the above embodiments can execute the short video popularity prediction method provided in any embodiment of this application. For technical details not described in detail in the above embodiments, please refer to the short video popularity prediction method provided in any embodiment of this application.

[0083] The above description is merely a preferred embodiment and the technical principles employed in this application. This application is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions that can be made by those skilled in the art will not depart from the scope of protection of this application. Therefore, although this application has been described in detail through the above embodiments, this application is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of this application. The scope of this application is determined by the scope of the claims.

Claims

1. A method for predicting the popularity of short videos, characterized in that, include: Obtain data information related to the short video to be predicted from multiple data sources; Based on the data information, the interactive feature information and text feature information of the short video to be predicted are determined; The popularity value of the short video to be predicted is obtained based on the interactive feature information and text feature information.

2. The short video popularity prediction method according to claim 1, characterized in that, The process of obtaining data information associated with the short video to be predicted from multiple data sources includes: Identify keywords related to the short video to be predicted, and expand these keywords to obtain a keyword set; Determine the interface corresponding to each of the data sources; Based on the keyword set, data information associated with the short video to be predicted is obtained from each of the data sources through the interface corresponding to each data source.

3. The short video popularity prediction method according to claim 2, characterized in that, The step of obtaining data information associated with the short video to be predicted from each of the data sources through the interface corresponding to each of the data sources based on the keyword set includes: Based on the keyword set and the interface corresponding to the data source, construct the request parameters corresponding to each data source; Based on the interface protocol of each data source and the request parameters, construct a query request to access the interface of each data source; The query request is sent to the corresponding interface of the data source, and the returned response data is received to obtain the data information associated with the short video to be predicted.

4. The short video popularity prediction method according to claim 1, characterized in that, The step of determining the interactive feature information and text feature information of the short video to be predicted based on the data information includes: The interactive data information corresponding to the short video to be predicted is obtained from the data information, and the interactive feature information of the short video to be predicted is determined based on the interactive data information, wherein the interactive data information includes any one or more of the following: number of likes, number of comments, number of plays, number of shares, and number of searches. The text data information corresponding to the short video to be predicted is obtained from the data information, wherein the text data information includes comment text information and bullet screen text information; Based on the comment text information and the bullet screen text information, the text feature information of the video to be predicted is determined.

5. The short video popularity prediction method according to claim 1, characterized in that, The process of predicting the popularity value of the short video to be predicted based on the interaction feature information and text feature information includes: The interaction feature information is quantized to obtain the interaction quantization value; Sentiment analysis is performed on the text feature information, and a sentiment quantification value is determined based on the analysis results; Based on the interaction quantification value and the emotion quantification value, the popularity value of the short video to be predicted is obtained.

6. The short video popularity prediction method according to claim 5, characterized in that, The step of performing sentiment analysis on the text feature information and determining a sentiment quantification value based on the analysis results includes: Sentiment analysis is performed on the text feature information to determine the sentiment tendency and sentiment intensity of the text feature information; The emotional quantification value is determined based on the emotional tendency and emotional intensity.

7. The short video popularity prediction method according to claim 5, characterized in that, The step of quantifying the interaction feature information to obtain the interaction quantization value includes: Each interaction dimension index in the interaction feature information is normalized to obtain the corresponding normalized value; The normalized values ​​corresponding to each of the interaction dimension indicators are weighted and summed to obtain the interaction quantification value.

8. A short video popularity prediction device, characterized in that, include: The information acquisition module is used to acquire data information related to the short video to be predicted from multiple data sources; The feature information determination module is used to determine the interactive feature information and text feature information of the short video to be predicted based on the data information; The popularity prediction module is used to predict the popularity value of the short video to be predicted based on the interaction feature information and text feature information.

9. A short video popularity prediction device, characterized in that, include: One or more processors; A memory that stores one or more programs, which, when executed by one or more processors, enable the one or more processors to implement the short video popularity prediction method as described in any one of claims 1-7.

10. A storage medium containing computer-executable instructions, characterized in that, The computer-executable instructions, when executed by a computer processor, are used to perform the short video popularity prediction method as described in any one of claims 1-7.

Images