KR20200139965A - System for video channel diagnostics and method thereof - Google Patents

Abstract

The present invention relates to a method for diagnosing a video channel, including: (a) a step of performing user authentication through a video content platform that provides a channel for sharing the video content among a plurality of user terminals, when channel diagnosis request information is received from the user terminal; (b) a step of performing scraping on at least one channel operated by the user terminal using the user authentication information to obtain raw channel data, when the user authentication is completed; (c) a step of processing the raw channel data into image content data obtained by extracting text information from the moving image content and user response data for the moving image content and storing the data in a database; (d) a step of calculating a channel analysis result for video content for each channel through a content analysis model which performs index analysis using the video content data and user response data; and (e) a step of providing channel growth guide data of content for inducing a user response for each channel based on the calculated channel analysis result.

Description

Video channel diagnosis system and its method {SYSTEM FOR VIDEO CHANNEL DIAGNOSTICS AND METHOD THEREOF}

The present invention relates to a video channel diagnosis system and method for performing objective analysis on the content of video content and user reaction.

Recently, web sites that provide video content services such as User Created Contents (UCC) through communication networks such as the Internet have been widely activated. In the past, mass media such as TV, newspaper, radio, and magazines that unilaterally delivered information to an unspecified number of people were dominated. However, with the development of information and communication technology, interactive communication is possible using the Internet and smartphones. With the advent of new media), single-person media is developing.

Here, one-person media is a new type of communication platform through which individuals can directly produce and share various contents as a medium for transmitting information, and refers to a new type of communication channel. As the market size of single-person video creators or creators grows, a new industrial field is emerging that is responsible for video production support and distribution, and generates revenue from various channels.

This one-person media industry is developing in the form of Multi-Channel Networks (MCN) and provides Internet-based video content platforms such as YouTube, Africa TV, and Twitch. In Korea, there are YouTube, Afreeca TV, Daum TV pod (or Kakao TV), and V Live, and there are many small Internet private broadcasters.

Creators are professions that create creative video content and upload it to video content platforms such as YouTube to secure subscribers and share advertising revenue with video content platform operators. These creators are eliciting explosive reactions to the 1020 generations by converting areas that were not covered or difficult to handle in terrestrial broadcasts, such as eating room (eating broadcast), game room (game broadcast), beauty, and cook room (cooking broadcast).

Video content platform operators are emerging as a new industrial form by discovering and supporting promising creators, or by collaborating with famous creators to create various advertisements and distributions. In recent years, it is expanding into business areas to create added value such as MD product production, online shopping mall opening, and entertainment.

As described above, as an environment in which anyone can freely create, distribute, and consume video content without restrictions, the number of creators and opening channels is increasing exponentially.

The biggest problem that can arise from creator-centered content production is the decline in the quality of the content or the reduction in the number of subscribers to the channel, resulting in restrictions on revenue generation.

However, training and training techniques for creators are currently being provided to produce video content, but objective and quantitative diagnosis has not been made on the channels that the creators have and the video content uploaded to each channel. Accordingly, creators other than famous creators cannot grasp the current status of their own channels or video content, and thus, it is difficult to improve the quality of content or generate revenue.

Korean Patent Registration No. 10-1429446 (Name of invention: Content creation system and method thereof)

In order to solve the above-described problem, according to an embodiment of the present invention, in accordance with an embodiment of the present invention, video contents uploaded to at least one channel held by a user at a request of a user terminal, as well as video contents, are index data indicating user reaction. The purpose is to provide the channel analysis results by objective and quantitative analysis using the channels.

In addition, an object of the present invention is to provide channel growth guide data for content planning, production, editing, exposure information, etc. to further induce user reaction based on the channel analysis result.

However, the technical problem to be achieved by the present embodiment is not limited to the technical problem as described above, and other technical problems may exist.

As a technical means for achieving the above technical problem, the video channel diagnosis method according to an embodiment of the present invention is performed by a video channel diagnosis system that analyzes video content distributed on at least one channel by a user terminal. A method for diagnosing a video channel, the method comprising the steps of: a) when receiving channel diagnosis request information from the user terminal, performing user authentication through a video content platform that provides a channel for sharing the video content between a plurality of user terminals; b) when the user authentication is complete, performing scraping on at least one channel operated by the user terminal using the user authentication information to obtain raw channel data; c) generating video content data obtained by extracting text information from the video content included in the raw channel data, user response data including inquiry information, exposure information, and viewing pattern information for the video content and storing it in a database; d) calculating a channel analysis result for video content for each channel through a content analysis model that performs index analysis using the video content data and user response data; And e) providing channel growth guide data including planning, production, editing, and exposure information for inducing user reactions for each channel based on the calculated channel analysis result, wherein the content analysis model comprises: Content classification is performed on video content data, and a correlation between the video content data and user response data is analyzed to provide a channel analysis result for a viewing pattern according to a user response for each video content.

In addition, a video channel diagnosis system according to another embodiment of the present invention is a program for performing a video channel diagnosis method in a video channel diagnosis system that analyzes video content distributed on at least one channel by a user terminal This recorded memory; And a processor for executing the program, wherein, when channel diagnosis request information is received from the user terminal by execution of the program, the processor provides a channel for sharing the video content between a plurality of user terminals. User authentication is performed through the provided video content platform, and when the user authentication is completed, scraping is performed on at least one channel managed by the user terminal using the user authentication information, and raw channel data ( Raw channel data), extracting text information from the video content included in the raw channel data, and generating user response data including inquiry information, exposure information, and viewing pattern information for the video content Stored in the image content data and user response data, the channel analysis result of the video content for each channel is calculated through a content analysis model that performs indicator analysis using the image content data and user response data, and induces a user response based on the calculated channel analysis result Provides channel growth guide data including planning, production, editing, and exposure information for content, wherein the content analysis model performs content classification on the video content data, and correlation between the video content data and user response data By analyzing the relationship, a channel analysis result for the viewing pattern according to the user reaction for each video content is provided.

According to the above-described problem solving means of the present invention, a user checks a lack of video content uploaded to at least one channel operated by the user, and objectively compares it with video content of other channels in the video content platform. You can statistically analyze your current channel situation. In addition, the user can use the channel growth guide data to improve the quality of the content and thereby create revenue by increasing the number of viewers.

According to the present invention, the quality of the content may be evaluated by comparing the video content for the user who requested the channel diagnosis with other video content in the video content platform, and the reaction of viewers may be compared and analyzed.

1 is a diagram illustrating the overall configuration of a video channel diagnosis system according to an embodiment of the present invention.
2 is a diagram illustrating a detailed configuration of a video channel diagnosis system according to an embodiment of the present invention.
3 is a diagram illustrating a program for performing a video channel diagnosis method according to an embodiment of the present invention.
4 is a flowchart illustrating a method of diagnosing a video channel according to an embodiment of the present invention.
5 is an exemplary diagram illustrating image content data according to an embodiment of the present invention.
6 is an exemplary diagram illustrating a word tokenization process according to an embodiment of the present invention.
7 is an exemplary diagram illustrating a word document matrix using a corpus of image content data according to an embodiment of the present invention.
8 is an exemplary diagram illustrating a content analysis model according to an embodiment of the present invention.
9 is an exemplary diagram for a channel analysis result according to an embodiment of the present invention.
10 is an exemplary diagram for providing guide contents of channel growth guide data according to an embodiment of the present invention.
11 and 12 are exemplary diagrams of channel growth guide data according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art can easily implement the present invention. However, the present invention may be implemented in various different forms and is not limited to the embodiments described herein. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and similar reference numerals are assigned to similar parts throughout the specification.

Throughout the specification, when a part is said to be "connected" to another part, this includes not only "directly connected" but also "electrically connected" with another element interposed therebetween. . In addition, when a part "includes" a certain component, it means that other components may be further included, and one or more other features, not excluding other components, unless specifically stated to the contrary. It is to be understood that it does not preclude the presence or addition of any number, step, action, component, part, or combination thereof.

In the present specification, a'terminal' may be a wireless communication device with guaranteed portability and mobility, and may be any kind of handheld-based wireless communication device, such as a smartphone, a tablet PC, or a notebook. In addition, the'terminal' may be a wired communication device such as a PC that can access other terminals or servers through a network. In addition, a network refers to a connection structure in which information exchange is possible between nodes such as terminals and servers, and includes a local area network (LAN), a wide area network (WAN), and the Internet (WWW). : World Wide Web), wired and wireless data networks, telephone networks, wired and wireless television networks, etc.

Examples of wireless data networks include 3G, 4G, 5G, 3rd Generation Partnership Project (3GPP), Long Term Evolution (LTE), World Interoperability for Microwave Access (WIMAX), Wi-Fi, Bluetooth communication, infrared communication, and ultrasound. Communication, Visible Light Communication (VLC), LiFi, etc. are included, but are not limited thereto.

The following examples are detailed descriptions to aid understanding of the present invention, and do not limit the scope of the present invention. Accordingly, the invention of the same scope performing the same function as the present invention will also belong to the scope of the present invention.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a view for explaining the overall configuration of a video channel diagnosis system according to an embodiment of the present invention, Figure 2 is a diagram illustrating a detailed configuration of a video channel diagnosis system according to an embodiment of the present invention, and FIG. Is a diagram illustrating a program for performing a video channel diagnosis method according to an embodiment of the present invention.

1 to 3, a video channel diagnosis system 100 includes a communication module 110, a memory 120, a processor 130, a database 140, and a display module 150.

The communication module 110 interworks with a communication network to provide a communication interface necessary for the video channel diagnosis system 100 to provide a transmission/reception signal between the user terminal 10 and the video content platform 200 in the form of packet data. Furthermore, the communication module 110 may perform a role of receiving a data request from a user terminal and transmitting data as a response thereto. Also, the communication module 110 may transmit a data request to the video content platform 200 and receive data as a response thereto.

Here, the communication module 110 may be a device including hardware and software necessary for transmitting and receiving a signal such as a control signal or a data signal through a wired or wireless connection with another network device.

On the other hand, the user terminal 10 is a terminal possessed by a single video creator or creator, and the video content platform 200 is a system providing a multi-channel network (MCN) service or YouTube provided by a business operator, It could be African TV, Twitch, Instagram, etc.

In the memory 120, a program for performing a video channel diagnosis method is recorded. Also, the memory 120 performs a function of temporarily or permanently storing data processed by the processor 130. Here, the memory 120 may include a volatile storage medium or a non-volatile storage medium, but the scope of the present invention is not limited thereto.

The processor 130 controls the entire process of providing the analysis result of the channel for sharing the video content to the user terminal 10 by executing a program for performing the video channel diagnosis method. Each operation performed by the processor 130 will be described in more detail later.

The video channel diagnosis system 100 may include at least one program module to perform at least one task. That is, as shown in FIG. 3, a program 300 for performing a video channel diagnosis method includes a collection module 310, an analysis module 320, and a visualization module 330.

The collection module 310 collects user authentication information provided by the video content platform 200 using an application programming interface (API), and raw channel data extracted through scraping for each channel. The raw channel data is stored in the database 240 through a pre-processing process.

The analysis module 320 performs analysis between indicators and analysis by indicators through a content analysis model, and provides an analysis result by analyzing a correlation between video content data of video content and user response data.

The visualization module 330 calculates values necessary to visualize the channel analysis result according to the channel analysis request information of the user terminal 10 or directly renders a graph or the like. If the user terminal 10 renders the channel analysis result using web technology, etc., the visualization module 330 responds with Asynchronous JavaScript and XML (Ajax) to only the channel analysis result data necessary for visualization. 10) so that the visualization is performed directly in the user terminal 10.

The above-described modules are only one embodiment for explaining the present invention, and are not limited thereto and may be implemented in various modifications. In addition, the above-described modules are stored in the memory 120 as a computer-readable recording medium that can be controlled by the processor 130. In addition, at least a portion of the program 300 may be implemented as software, firmware, hardware, or a combination of at least two or more thereof, and may include a module, program, routine, instruction set, or process for performing one or more functions. .

The processor 130 may include all kinds of devices capable of processing data, such as a processor. Here, the'processor' may refer to a data processing device embedded in hardware having a circuit physically structured to perform a function represented by a code or instruction included in a program. As an example of a data processing device built into the hardware as described above, a microprocessor, a central processing unit (CPU), a processor core, a multiprocessor, and an application-specific integrated (ASIC) circuit) and processing devices such as field programmable gate arrays (FPGAs), but the scope of the present invention is not limited thereto.

The database 140 stores data accumulated while performing a video channel diagnosis method. For example, the database 140 may store user authentication information, raw channel information, image content data, user response data, channel analysis result data, and the like.

The display module 150 displays a text, table, or graph report on the channel analysis result data according to the channel diagnosis request information of the user terminal 10 under the control of the processor 130.

4 is a flowchart illustrating a method of diagnosing a video channel according to an embodiment of the present invention.

The user terminal 10 creates at least one personal channel among a plurality of channels operated by the video content platform 200 and uploads and operates the video content.

When receiving the channel diagnosis request information from the user terminal 10 (S1), the video channel diagnosis system 100 requests user authentication information such as ID/password, phone number, and login information from the user terminal 10 (S2). ).

When the user authentication information is received from the user terminal 10 (S3), the video channel diagnosis system 100 checks the validity or validity of the user authentication information through the video content platform 200 (S4). At this time, the video channel diagnosis system 100 proceeds with a user consent procedure through the user terminal 10 to receive user authentication information from the video content platform 200 to which the user subscribes or belongs.

In addition, the video channel diagnosis system 100 may store user authentication information in the database 140 in the case of a user who has undergone a user authentication procedure one or more times, and may retrieve and use user authentication information from the database at the time of next user authentication. .

When user authentication for the user terminal 10 fails within a preset number of times, an authentication failure message or an authentication retry request message is transmitted to the user terminal 10 (S5, S6).

However, when the number of user authentication failures of the user terminal 10 exceeds the preset number, it is checked through the video content platform 200 whether it corresponds to blacklist information that is blocked from access due to harmful or illegal video content. When the user authentication information of the user terminal 10 is not blacklist information, the video channel diagnosis system 100 outputs an authentication information reconfirmation message for expiration or error input of the user authentication information. When the user authentication information corresponds to the blacklist information, a channel diagnosis request rejection message may be output for a preset period in connection with the video content platform 200.

On the other hand, the video channel diagnosis system 100 checks the user authentication information for the user terminal 10, and after the user authentication is normally completed, the raw channel data for the personal channel of the user terminal through the video content platform 200 Bring it (S7). That is, by using the user authentication information, the video content platform 200 scrapes for each channel to obtain and collect raw raw channel data.

The video channel diagnosis system 100 scrapes the video content platform 200 using user authentication information to obtain raw channel data, and performs a process of updating user authentication information when the scraping attempt fails.

A user, that is, a creator, may have a plurality of channels in the video content platform 200, and a plurality of video contents may exist for each channel. Accordingly, the video channel diagnosis system 100 fetches overall data for each channel and metadata for video contents existing in the channel.

The raw channel data includes video content data such as audio, video, image, metadata, title, description tag, and category. In addition, it includes various user reaction data such as video length, number of likes/dislikes, viewing time, number of viewers by video section, viewer retention rate, number of views, number of impressions, number of clicks versus exposure, and viewer demographic information. In this case, the image content data may include user response data such as positive words and negative words.

The video channel diagnosis system 100 calculates a channel analysis result by analyzing data for each channel or for each video content through a content analysis model based on the raw channel data or data preprocessed from the raw channel data (S8, S9).

In addition, the video channel diagnosis system 100 provides channel growth guide data for planning, production, editing, exposure, etc. of moving video content according to trends based on the channel analysis result as well as the channel analysis result (S10). The channel analysis result may include an age group analysis result for each viewer, a video content change analysis result for each time period, a video content data analysis with the highest click rate, and the like.

5 is an exemplary diagram illustrating image content data according to an embodiment of the present invention, and FIG. 6 is an exemplary diagram illustrating a word tokenization process according to an exemplary embodiment of the present invention.

The video channel diagnosis system 100 collects raw channel data, pre-processes it, and stores it in the database 140 as image content data and user response data.

In this case, the pre-processing process for the image content data may consist of a process of extracting words to analyze individual words and a process of tokenizing words for a content analysis model based on deep learning.

First, the process of extracting words to analyze individual words is a topic modeling algorithm including TF-IDF (Term Frequency-Inverse Document Frequency), LDA (Latent Dirichlet Allocation, LDA), or hLDA (Hierarchical LDA). By using, words with a preset frequency or more are extracted as text information from the image content data.

As an example, as shown in FIG. 5, nouns are extracted from image content data, and'Kokko (creator's unique name)','Tipping', and'Kokoya (subscriber nickname)' using a topic modeling algorithm such as TF-IDF. Extract important words such as'. Words extracted in this way are rarely used in other video content data, but they are heavily used in the video content data, so the weight is given high. In addition, statistical values for words such as a bag of words (BOW) can be used, and data on the frequency of words in the video content data such as 3'applications' and 5'today' are used.

On the other hand, the word tokenization process for a content analysis model based on deep learning removes noise from natural language data extracted from image content data, and sets words extracted through stemming or lemmatization. Text information is extracted by word tokenization based on.

Since Korean, foreign languages, foreign languages, special characters, and transformed words are widely used in natural language used in image content data, a refinement operation to remove noise and a normalization operation to integrate words of various expression methods are performed.

As shown in FIG. 6, word tokenization is performed through normalization and stem extraction of some sentences of image content data, and natural language analysis is performed using word tokens when tokenization is completed.

7 is an exemplary diagram illustrating a word document matrix using a corpus of image content data according to an exemplary embodiment of the present invention, and FIG. 8 is an exemplary diagram illustrating a content analysis model according to an exemplary embodiment of the present invention.

The content analysis model derives a channel analysis result for the viewing pattern according to the response level of each video content by analyzing the correlation between the video content data and the user response data, performing content classification on the video content data. Such a content analysis model can use a basic analysis method based on a regression model or a natural language analysis method using deep learning.

The basic analysis method based on the regression model collects topic words or corpus data extracted from image content data through a pre-processing process to generate a set of word frequency counts for each image content data (or document data), and For example, a Term Document Matrix (TDM) can be generated. Since the word document matrix is a high-dimensional matrix, a dimension reduction technique can be applied. After the word document matrix is passed through the TF-IDF algorithm, the similarity between images is obtained through the cosine similarity. In addition, a regression model to analyze the correlation between each image content data and an index is established, and a learning process for a plurality of image content data is performed through the regression model.

In this case, the similarity between images may be calculated using a similarity calculation method between various vectors such as an L1 distance (L1 Distance or Manhattan Distance) and an L2 distance (L2 Distance or Euclidean Distance) as well as cosine similarity.

Meanwhile, the natural language analysis method using deep learning is doc2vec's Paragraph Vector with Distributed Memory (PV-DM) or PV-DBOW (Paragraph Vector with Distributed Bag of), which is a document embedding method that converts image content data into vectors. words Memory) algorithm is used to calculate the image content vector.

As shown in FIG. 8, the natural language analysis method using deep learning can predict an index value for a large amount of video content through learning using a neural network model including a hidden layer composed of 1 to 3 layers.

In other words, after learning natural language data using a large amount of natural language corpus in the neural network model, it is possible to extract a vector containing semantic content for video content by additionally learning the video content data, and improve natural language understanding performance. I can. As described above, the natural language analysis method using deep learning can obtain a large number of word document matrices compared to the analysis method based on a regression model, and each word token using a natural language processing model such as BERT (Bidirectional Encoder Representations from Transformers) algorithm. You can perform an analysis considering the relationship between them.

A content analysis model using a regression model or a neural network model may select a category of each video content by performing content classification using video content data of video content for each channel.

For example, the categories of video content are ① TV Jalbang (shortly edited broadcast video), ② Educational contents such as various education and informational contents, ③ Celebrity, singer, performance related videos, ④ Meokbang/Cookbang, ⑤ Sports, ⑤ Game broadcasting. , ⑥ web drama, web entertainment, talk show, mobile video, ⑦ vlog, ⑧ news, ⑨ animals, ⑩ beauty, ⑪ unboxing, ⑫ product reviews such as user reviews, ⑬Kids, ⑭ASMR, etc.

The content analysis model uses raw channel data or video content data pre-processed from raw channel data and user response data as input data, and extracts user response data, processed data (or indicator data), label data, etc., extracted through scraping. You can do it with output data. Table 1 describes raw channel data and processed data.

Raw channel data Processed data (or indicator data) Number of likes, number of dislikes Like (positive)/dislike (negative) ratio Viewer retention rate compared to video length and video progress (unit: %) Percentage of viewers remaining compared to video duration (unit: seconds) Views, likes Likes (positive) ratio to views Audio data of video Text data for speech Video data Compressed video frame at 1FPS level

In particular, when label data is set as the output data of the content analysis model, the positive/negative reaction score by viewers who watched the video content by assigning the label data to user reaction data including positive/negative reactions, fun, and empathy, and fun. Through scoring including scores and empathy scores, it is possible to learn the label data for the corresponding video content, and evaluate the quality of the video content by using the learned label data.

9 is an exemplary diagram for a channel analysis result according to an embodiment of the present invention.

As shown in FIG. 9, the video channel diagnosis system 100 may analyze indicator data according to the type of video content for an individual channel, that is, a category. Here, the indicator data may be user response data such as user demographic information, viewing time, click-through rate, views, and likes (positive) ratio.

The channel analysis result may provide a result of analyzing the contents of the video content, an average viewing time for each category, an average video length, and a residual rate according to the viewing time in a graph form. The retention rate according to the viewing time is based on data such as the average viewing time compared to the total length of the video, so that the quality of the video content and the actual preferred video of the viewers can be analyzed.

For example, if the number of views of food content among video contents uploaded to Creator B's personal channel is greater than the number of views of daily content, the channel analysis result includes a positive evaluation of the viewing pattern for food content. May establish a strategy of uploading food content for growth of its own channel through correlation between food content and indicator data (ie, user response data) based on the channel analysis result.

Alternatively, the channel analysis result may include a result of analyzing a change trend of user reaction data for each video content for each individual channel in a time series. Accordingly, when the video content is periodically uploaded to the creator B's personal channel, the creator B can determine which content has the highest recent user response data (eg, views, etc.) through the channel analysis result.

The channel analysis result may be provided by analyzing viewer demographic data for each video content to select and provide a video content having a maximum responsiveness of user response data by age or gender. Accordingly, the creator B may select a target target of the gender or age group he/she wants to target, check the video content matching the target target based on the channel analysis result, and apply it to image planning and production.

10 is an exemplary diagram for providing guide content of channel growth guide data according to an embodiment of the present invention, and FIGS. 11 and 12 are exemplary diagrams for channel growth guide data according to an embodiment of the present invention.

As shown in FIG. 10, the channel growth guide data may provide a guideline including video planning, production, editing, and exposure information (e.g., thumbnails, titles, etc.) necessary for channel growth based on the channel analysis result. have.

The video channel diagnosis system 100 calculates an image similarity between the learned channel analysis result of the video contents of other channels learned through the content analysis model and the channel analysis result of the video content corresponding to the personal channel of the user terminal 10. , Channel growth guide data including planning, production, editing, and exposure information of the video content may be provided through comparative learning using the extracted similar video content after extracting the highest similar video content among the calculated image similarities.

As shown in FIG. 10, when it is confirmed that the video content of'Chinese food eating' and'dog' is uploaded through the channel analysis result of creator A, the channel growth guide data is Chinese food eating among similar video contents. The user's reaction can be analyzed by exposing popular videos of content and popular videos of dog content to guidelines.

At this time, the images exposed to the guideline are referred to as guide content, and at least one guide content may be provided for each category of video content. In this case, guide content is analyzed through trend analysis using user response data for similar video content. Select.

Meanwhile, the popularity order of similar video contents is determined using a popularity index including views, viewing time, and like (positive) ratio, and guide contents are arranged in the determined popularity order and provided as popular videos.

As shown in FIGS. 11 and 12, the channel growth guide data includes information on inducing user reactions according to the target of the channel, information on the video immersion time according to the average viewing time of the video, and the current situation of viewer retention and departure by video playback time It contains elements necessary for channel growth and comparison information with popular video contents of other channels. In addition, content element information for increasing the number of viewers according to the early, mid, and late regions of the video content is provided in the tip part.

Therefore, Creator 1 can receive recommendations for thumbnails and titles for inducing user reactions through channel growth guide data, and can objectively identify insufficient parts of his or her video content through comparative learning with similar video contents. .

The video channel diagnosis method according to the embodiment of the present invention described above may be implemented in the form of a recording medium including instructions executable by a computer such as a program module executed by a computer. Such recording media include computer-readable media, and computer-readable media may be any available media that can be accessed by a computer, and include both volatile and nonvolatile media, and removable and non-removable media. In addition, computer-readable media includes computer storage media, which are volatile and nonvolatile embodied in any method or technology for storage of information such as computer-readable instructions, data structures, program modules or other data. , Removable and non-removable media.

The above description of the present invention is for illustrative purposes only, and those of ordinary skill in the art to which the present invention pertains will be able to understand that other specific forms can be easily modified without changing the technical spirit or essential features of the present invention will be. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not limiting. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as being distributed may also be implemented in a combined form.

The scope of the present invention is indicated by the claims to be described later rather than the detailed description, and all changes or modified forms derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention. do.

100: video channel diagnosis system
110: communication module 120: memory
130: processor 140: database
150: display module

Claims (22)

In the video channel diagnosis method performed by a video channel diagnosis system for analyzing video content distributed on at least one channel by a user terminal,
a) when channel diagnosis request information is received from the user terminal, performing user authentication through a video content platform that provides a channel for sharing the video content between a plurality of user terminals;
b) when the user authentication is complete, performing scraping on at least one channel operated by the user terminal using the user authentication information to obtain raw channel data;
c) generating video content data obtained by extracting text information from the video content included in the raw channel data, user response data including inquiry information, exposure information, and viewing pattern information for the video content and storing it in a database;
d) calculating a channel analysis result for video content for each channel through a content analysis model that performs index analysis using the video content data and user response data; And
e) comprising the step of providing channel growth guide data including planning, production, editing, and exposure information for inducing user reactions for each channel based on the calculated channel analysis result,
The content analysis model performs content classification on the video content data, analyzes a correlation between the video content data and user response data, and provides a channel analysis result for a viewing pattern according to a user response for each video content. That, video channel diagnosis method. The method of claim 1,
The step a),
a-1) If user authentication for the user terminal fails within a preset number of times, an authentication failure message or an authentication retry request message is output,
a-2) If the number of user authentication failures of the user terminal exceeds a preset number, confirming whether it corresponds to blacklist information that is blocked from access related to harmful or illegal video content through the video content platform. That includes, video channel diagnosis method. The method of claim 2,
Step a-2)
If the user authentication information of the user terminal is not blacklist information, an authentication information reconfirmation message for the expiration of the user authentication information or error input is output, and
When the user authentication information of the user terminal corresponds to blacklist information, outputting a channel diagnosis request rejection message for a predetermined period in connection with the video content platform. The method of claim 1,
Step b),
If the scraping has failed, updating the user authentication information, and requesting scraping from the video content platform using the updated user authentication information. The method of claim 1,
Step c)
Extracting text information including video, audio, metadata, title, description, tag, comment, or category from the video content to generate the video content data,
To generate user response data representing user reactions, including positive/negative ratings, views, impressions, clicks versus impressions, viewer retention rate, viewing time, or viewer demographic information for the video content. Way. The method of claim 1,
Step c)
Text information of words with a preset frequency or more in the video content using a topic modeling algorithm including TF-IDF (Term Frequency-Inverse Document Frequency), LDA (Latent Dirichlet Allocation, LDA), or hLDA (Hierarchical LDA). That is, the video channel diagnosis method. The method of claim 1,
Step c)
Text information is extracted by removing noise from the natural language data extracted from the video content, and performing word tokenization based on preset rules on words extracted through stemming or Lemmatization. That is, how to diagnose a video channel. The method of claim 1,
Step d)
In the content analysis model, using raw channel data of a plurality of video content or natural language word information in the video content data to learn indicator data associated with the user response data,
After classifying a content category using video content data of video content for each channel through the learned content analysis model, based on the correlation analysis result between the video content data of the classified content and user response data, the The method for diagnosing a video channel to select a content category having a maximum value of user response data. The method of claim 1,
Step d)
In the content analysis model, using raw channel data of a plurality of video content or natural language word information in the video content data to learn indicator data associated with the user response data,
A video channel that analyzes a change trend of user response data for each video content for each channel through the learned content analysis model in a time series, and selects a time zone and video content having a maximum responsiveness of the user response data Diagnostic method. The method of claim 1,
Step d)
In the content analysis model, using raw channel data of a plurality of video content or natural language word information in the video content data to learn indicator data associated with the user response data,
By analyzing viewer demographic data for each video content through the learned content analysis model, video content having a maximum reactivity of user response data by age or gender is selected. The method of claim 1,
Step e)
The image similarity between the channel analysis result of the video content of the other channel learned through the content analysis model and the channel analysis result of the video content corresponding to the channel of the user terminal is calculated, and the highest similarity among the calculated image similarity To extract video contents and provide channel growth guide data including planning, production, editing, and exposure information of video contents through comparative learning using the extracted similar video contents. The method of claim 11,
The image similarity is calculated by using any one of cosine similarity, L1 distance (L1 Distance or Manhattan Distance), and L2 distance (L2 Distance or Euclidean Distance). The method of claim 11,
Step e)
A method for diagnosing a video channel by providing at least one video content among the similar video content as guide content, and providing the guide content for each category of the video content. The method of claim 13,
A method for diagnosing a video channel, wherein guide content is selected through trend analysis using user response data for the similar video content. The method of claim 11,
Step e)
A popularity order is determined by using a popularity index including views, viewing time, and affirmative ratio among user reaction data for the similar video contents, and the similar video contents are arranged in the determined popularity order, and video channel diagnosis Way. The method of claim 1,
Step e),
Label data is assigned to user reaction data including positive/negative reactions, fun, and empathy for the video content, and score including positive/negative reaction scores, fun scores, and empathy scores by viewers who watched the video content A method for diagnosing a video channel, which is to learn as label data for video content through and evaluate the video content using the learned label data. In the video channel diagnosis system for analyzing video content distributed on at least one channel by a user terminal,
A memory in which a program for performing a video channel diagnosis method is recorded; And
And a processor for executing the program,
The processor, by executing the program,
When channel diagnosis request information is received from the user terminal, user authentication is performed through a video content platform that provides a channel for sharing the video content between a plurality of user terminals,
When the user authentication is completed, scraping is performed on at least one channel managed by the user terminal using the user authentication information to obtain raw channel data,
Generated as image content data extracted from the text information from the video content included in the raw channel data and user response data including inquiry information, exposure information, and viewing pattern information for the video content, and stored in a database,
Calculate a channel analysis result for video content for each channel through a content analysis model that performs index analysis using the video content data and user response data,
Provide channel growth guide data including planning, production, editing, and exposure information for content planning, production, editing, and exposure information for inducing user reaction based on the calculated channel analysis result,
The content analysis model performs content classification on the video content data, analyzes a correlation between the video content data and user response data, and provides a channel analysis result for a viewing pattern according to a user response for each video content. That, video channel diagnosis system. The method of claim 17,
The video channel diagnosis system further comprising a display module for outputting a channel analysis result report in the form of a text, a table, or a graph under the control of the processor. The method of claim 18,
The processor,
Computing visualization values of the channel analysis result, rendering the channel analysis result report, and outputting it to the display module. The method of claim 18,
The processor,
A video channel diagnosis system to extract the channel analysis result data required for visualization in response to Asynchronous JavaScript and XML (Ajax) so that the channel analysis result report is displayed on the user terminal and transmit it to the user terminal. The method of claim 17,
The processor is
Using API (Application Programming Interface), user authentication information and raw channel data through channel-specific scraping are imported through the video content platform,
When the scraping fails, the user authentication information is updated and then the video content platform is requested to scrape using the updated user authentication information. The method of claim 17,
The channel growth guide data,
Based on the channel analysis result, at least one of information on inducing user reactions according to targeting targets, information on video immersion time according to average video viewing time, information on viewer retention and departure by video playback time slot, or comparison information with video content of other channels That includes, video channel diagnosis system.

Images