Electronic apparatus and control method thereof

Abstract

The electronic device includes a processor that acquires information on weights assigned to each of a plurality of items having values ​​representing content characteristics, wherein the weights are assigned higher to items among the plurality of items that correspond to content characteristics preferred by multiple users by content genre, identifies content viewed by a user based on the viewing history of the user of the electronic device, calculates a recommendation score for each of the plurality of contents based on the similarity of the values ​​of each item between the multiple recommendable contents and the identified content, and the weights assigned to each item by genre, and performs a recommendation operation regarding the content with the high calculated recommendation score among the multiple recommendable contents.

Description

Technology Field

[0001] The present invention relates to an electronic device that displays a content image by playing content selected from among several provided contents, and a control method thereof. More specifically, it relates to an electronic device that selects content suitable for a user from among several provided contents and recommends it to the user, and a control method thereof. Background Technology

[0002] Electronic devices, which fundamentally include electronic components such as CPUs, chipsets, and memory for computation to perform calculations and process specific information according to a particular process, can be classified into various types depending on the information to be processed or the intended use. For example, electronic devices include information processing devices such as PCs or servers that process general-purpose information, image processing devices that process image data, audio devices that process audio, and household appliances that perform chores within the home. An image processing device can be implemented as a display device that displays processed image data as an image on a display panel equipped with itself. The electronic device provides a UI that allows the user to select specific content from among numerous contents provided by external devices such as servers or stored internally, and plays the content selected through the UI to display the content image.

[0003] As content provision services expand and the number of contents available from various content providers increases, the issue of which content to recommend to the user from among the diverse and multiple available contents (hereinafter referred to as "valid content") becomes a concern. For example, it is practically difficult for an electronic device to display selection options for all valid content within its UI. Therefore, the electronic device selects multiple contents from the valid content that are expected to be highly preferred by the user and displays a UI that includes selection options for these selected contents. By doing so, the electronic device can go beyond simply presenting selectable options and more actively enhance user convenience regarding content selection by guiding the user toward recommended content they prefer.

[0004] Various methods have been proposed to determine which content among valid content has high user preference. However, with the volume of provided content increasing rapidly, these proposed methods sometimes result in selecting content that does not match the user's tendencies as recommended content. For example, there is a method that prioritizes recommending channels with high viewing times based on the user's viewing history. While this method can select channels or programs that the user has frequently watched, it cannot select channels or programs that the user has no viewing history of as recommended content.

[0005] Therefore, electronic devices may require a method for selecting recommended content that more accurately estimates and reflects user preferences among valid content (i.e., more accurately matches the user's tendencies). means of solving the problem

[0006] An electronic device according to an embodiment of the present invention includes a processor that acquires information on weights assigned to each of a plurality of items having values ​​representing content characteristics, wherein the weights are assigned highly to items corresponding to content characteristics preferred by a plurality of users according to the genre of content among the plurality of items, identifies content viewed by a user based on the viewing history of a user of the electronic device, calculates a recommendation score for each of the plurality of content based on the similarity of values ​​of each of the items between a plurality of recommendable content and the identified content, and the weights assigned to each of the items according to the genre, and performs a recommendation operation regarding content with a high calculated recommendation score among the plurality of recommendable content.

[0007] In addition, the processor may calculate item-specific similarity by mutually comparing the item-specific values ​​between a predetermined first content and a second content identified as having been viewed by the user among the plurality of contents, identify the weights corresponding to each item, and calculate the total value obtained by reflecting each identified weight in the item-specific similarity as the recommendation score of the first content.

[0008] In addition, the value of the first item among the plurality of items is represented as a numerical value, and the processor can identify a higher degree of similarity for each first item with respect to the first content as the difference between the first content and the second content is smaller.

[0009] In addition, the value of the second item among the plurality of items is displayed as text, and the processor can identify a higher degree of similarity for each second item to the first content as the number of common keywords between the values ​​of the second item of each of the first content and the second content increases.

[0010] In addition, the processor can identify the content with the highest viewing frequency among the content included in the viewing history as the second content.

[0011] In addition, the processor can identify the genre with the highest viewing frequency among the content included in the viewing history and identify the content of the identified genre as the second content.

[0012] In addition, the processor may display a UI on a display that enables the selection of one or more contents having a recommendation score higher than a preset rank among the plurality of contents.

[0013] In addition, the processor can identify a plurality of recommendable contents based on EPG (electronic program guide) information.

[0014] In addition, a control method for an electronic device according to an embodiment of the present invention comprises: a step of obtaining information on weights assigned to each of a plurality of items having values ​​representing content characteristics; a step of assigning a high weight to an item among the plurality of items that corresponds to content characteristics preferred by a plurality of users by genre of content, and a step of identifying content viewed by a user based on the viewing history of a user of the electronic device; a step of calculating a recommendation score for each of the plurality of contents based on the similarity of the values ​​of each item between a plurality of recommendable contents and the identified content and the weight assigned to each item by genre; and a step of performing a recommendation operation regarding the content with a high calculated recommendation score among the plurality of recommendable contents. Brief explanation of the drawing

[0015] Figure 1 is a block diagram of an electronic device. Figure 2 is an example of an electronic device displaying a UI that guides recommended content. Figure 3 is a flowchart regarding how an electronic device identifies recommended content. Figure 4 is a block diagram showing the schematic operation of an electronic device. Figure 5 is an example diagram showing the data structure of each content. Figure 6 is a table showing the characteristics of content genre preferences and the items corresponding to each characteristic. Figure 7 is a table showing examples of weights assigned to each item by genre. Figure 8 is a table showing an example of a user's viewing history of an electronic device. Figure 9 is an example diagram showing a method for calculating item-by-item similarity between recommendable content and viewed content. Figure 10 is an example diagram showing a method for calculating a recommendation score for the content by reflecting weights in the calculated similarity. Figure 11 is an example of an electronic device displaying a UI for genre selection. Specific details for implementing the invention

[0016] Embodiments according to the present invention will be described in detail below with reference to the accompanying drawings. Unless otherwise noted, the embodiments described with reference to each drawing are not mutually exclusive and multiple embodiments may be selectively combined and implemented within a single device. Such combinations of multiple embodiments may be arbitrarily selected and applied by a person skilled in the art to implement the concept of the present invention.

[0017] If, in the embodiments, terms including ordinal numbers such as "first component," "second component," etc. are used, such terms are used to describe various components and to distinguish one component from another; therefore, the meaning of these components is not limited by the terms. The terms used in the embodiments are applied to describe the corresponding embodiments and do not limit the scope of the invention.

[0018] Additionally, where the expression "at least one" appears among the plurality of components in this specification, this expression refers not only to all of the plurality of components but also to each individual component excluding the rest, or any combination thereof.

[0019] Figure 1 is a block diagram of an electronic device.

[0020] As illustrated in FIG. 1, the electronic device (100) according to the present embodiment can play content provided from various external devices and provide it to the user. The electronic device (100) can perform the operation according to the present embodiment, which will be described later, independently. Alternatively, depending on the design method, the electronic device (100) may perform the operation according to the present embodiment in conjunction with a communicable server (200).

[0021] The electronic device (100) can be implemented in various types, such as a fixed display device including a TV, monitor, digital signage, electronic whiteboard, electronic picture frame, etc.; an image processing device including a set-top box, optical multimedia player, etc.; an information processing device including a computer main body, etc.; a mobile device including a smartphone, tablet device, etc.; an imaging device including a camera, camcorder, etc.; a home appliance including a washing machine, refrigerator, clothing care device, air conditioner, electric vacuum cleaner, etc.; a wearable device provided for a user to wear on their body; or a communication device such as a hub, gateway, router, AP, etc.

[0022] The electronic device (100) and the server (200) include various hardware elements for operation. In this embodiment, the case where the electronic device (100) is a TV is described. However, since the electronic device (100) can be implemented as various types of devices as previously described, the configuration described below may vary depending on the type of electronic device (100) implemented.

[0023] The electronic device (100) may include an interface unit (110). The interface unit (110) includes an interface circuit for the electronic device (100) to communicate with various external devices, including a server (200), and to transmit and receive data. Depending on the connection method, the interface unit (110) includes at least one of one or more wired interface units (111) for wired communication connection or one or more wireless interface units (112) for wireless communication connection.

[0024] The wired interface section (111) includes a connector or port to which a cable of a predefined transmission standard is connected. For example, the wired interface section (111) includes a port to which a cable of a cable broadcasting system is connected to a terrestrial or satellite broadcasting antenna to receive a broadcast signal. Alternatively, the wired interface section (111) includes a port to which a cable of various wired transmission standards, such as HDMI, DP (DisplayPort), DVI, component, composite, S-Video, and Thunderbolt, is connected to connect to various video processing devices. Alternatively, the wired interface section (111) includes a USB standard port to connect to a USB device. Alternatively, the wired interface section (111) includes an optical port to which an optical cable is connected. Alternatively, the wired interface section (111) includes an audio input port to which an external microphone is connected, and an audio output port to which a headset, earphones, external speakers, etc. are connected. Alternatively, the wired interface section (111) includes an Ethernet port for connecting to a gateway, router, hub, etc., to connect to a wide area network.

[0025] The wireless interface section (112) includes a bidirectional communication circuit comprising at least one of the components, such as a communication module and a communication chip, corresponding to various types of wireless communication protocols. For example, the wireless interface section (112) includes a Wi-Fi communication chip that performs wireless communication with an AP (Access Point) according to the Wi-Fi method, a communication chip that performs wireless communication such as Bluetooth, Zigbee, Z-Wave, WirelessHD, WiGig, NFC, an IR module for IR communication, and a mobile communication chip that performs mobile communication with a mobile device.

[0026] The electronic device (100) may include a display unit (120). The display unit (120) forms a screen for displaying an image signal processed by a processor (170) as an image. The display unit (120) includes a display panel, and various design methods may be applied to the structure of the display panel. For example, the display unit (120) may include a light-receiving structure display panel, such as a liquid crystal, and a backlight that provides light thereto. Alternatively, the display unit (120) may include a self-emissive structure display panel, such as an OLED. Alternatively, the display unit (120) may have a structure in which a plurality of micro LED modules are combined in a tile form to form a large screen.

[0027] The electronic device (100) may include a user input unit (130). The user input unit (130) includes various types of user input interface-related circuits provided to allow the user to operate in order to perform user input. The user input unit (130) may be configured in various forms depending on the type of electronic device (100), and may include, for example, a mechanical button unit or an electronic button unit of the electronic device (100), various types of sensors, a touchpad, a touchscreen installed on a display, etc. Alternatively, the user input unit (130) may include external input devices such as a keyboard, mouse, or remote controller that are separated from the electronic device (100) and connected through an interface unit (110). Alternatively, the electronic device (100) may receive a control signal corresponding to user input from an external device, such as a mobile device, through the interface unit (110), in addition to the user input unit (130).

[0028] The electronic device (100) may include a storage unit (140). The storage unit (140) stores digitized data. The storage unit (140) includes storage with non-volatile properties that can preserve data regardless of whether power is provided, and memory with volatile properties in which data to be processed by the processor (170) is loaded and data cannot be preserved if power is not provided. Storage includes flash memory, hard-disc drive (HDD), solid-state drive (SSD), and ROM (Read Only Memory), and memory includes buffer and RAM (Random Access Memory).

[0029] The electronic device (100) may include a speaker (150). When the processor (170) plays a predetermined content, the speaker (150) outputs an audio signal of the content as sound. The speaker (150) may be installed in the electronic device (100) or provided as a separate device. If the speaker (150) is provided as a separate device, the speaker (150) is connected to an interface unit (110), and the audio signal is transmitted to the speaker (150) through the interface unit (110).

[0030] The electronic device (100) may include a microphone or a microphone set (160). The microphone set (160) collects sounds generated in the surrounding environment of the electronic device (100) to generate an audio signal and transmits this audio signal to a processor (170). In this embodiment, the microphone set (160) is described as a separate component from the user input unit (130), but depending on the design method, the microphone set (160) may be considered as a component included in the user input unit (130).

[0031] The electronic device (100) may include a processor (170). The processor (170) includes one or more hardware processors, such as a CPU, chipset, buffer, and circuit, which are mounted on a printed circuit board, and may be implemented as an SOC depending on the design method. When the electronic device (100) is a display device, the processor (170) includes modules corresponding to various processes, such as a demultiplexer, decoder, scaler, audio DSP (Digital Signal Processor), and amplifier, to display video content as video. Here, some or all of these modules may be implemented as an SOC. For example, modules related to video processing, such as a demultiplexer, decoder, and scaler, may be implemented as a video processing SOC, and the audio DSP may be implemented as a chipset separate from the SOC. The processor (170) plays a predetermined content so that the image of the content is displayed on the display unit (120), while the audio of the content is output as sound through the speaker (150).

[0032] The server (200) also includes hardware elements such as an interface unit (210), a wired interface unit (211), a wireless interface unit (212), a user input unit (230), a storage unit (240), and a processor (270). Since these components of the server (200) are basically similar to the components of the same name described earlier in the electronic device (100), a detailed description is omitted.

[0033] Below, a method for an electronic device (100) to guide a user to recommended content is described.

[0034] Figure 2 is an example of an electronic device displaying a UI that guides recommended content.

[0035] As illustrated in FIGS. 1 and 2, the electronic device (100) can receive various content from various content providing devices capable of communicating through the interface unit (110). The content providing device is not limited in type as long as it is a device capable of providing a predetermined content to the electronic device (100), and may include, for example, a server (200) providing streaming, VOD (Video On Demand), or OTT (Over The Top) services, broadcasting equipment of a broadcasting station, a set-top box, an optical media player, a mobile device, a computer main body, etc. Among the various content provided from the content providing device in this manner, the electronic device (100) selects one or more recommended content that matches the preferences of the currently identified user (for example, the user of the currently logged-in account) and provides them to the user. As a method of providing recommended content, the electronic device (100) displays a UI (10) on the display unit (120).

[0036] The UI (10) includes a menu (11) for one or more recommended content. Each content menu (11) displays information about one or more recommended content (e.g., a thumbnail image of a scene of the content, title, production company, playback time, basic information, etc.) and provides selection options for the recommended content. When a user selects a menu (11), the electronic device (100) switches to a detailed information page for the content of the selected menu (11), switches to a purchase page for the content if the content is subject to purchase, displays control options related to the playback of the content, or plays the content.

[0037] When the number of menus (11) included in the UI (10) is so large that it extends beyond the screen of the display unit (120), the electronic device (100) displays a scroll bar (12) on one side of the display unit (120) so that scrolling of the UI (10) using the scroll bar (12) is possible.

[0038] The electronic device (100) identifies one or more pieces of content that match the user's preferences as recommended content among many pieces of content that can be obtained. Below, a method for the electronic device (100) to identify one or more pieces of recommended content among the available content will be described.

[0039] Figure 3 is a flowchart regarding how an electronic device identifies recommended content.

[0040] As illustrated in FIGS. 1 and FIGS. 3, the following operation is performed by the processor (170) of the electronic device (100).

[0041] In step 310, the electronic device (100) obtains information on weights assigned to multiple items having values ​​representing content characteristics. Here, among the multiple items, a high weight is assigned to the item corresponding to the content characteristics preferred by multiple users according to the genre of the content. The multiple users referred to in this step refer to, for example, users of various clients connected to communicate with the server (200). The multiple users may or may not include users of the electronic device (100).

[0042] A given piece of content possesses metadata, and the characteristics of that content are represented by the values ​​of multiple items included in the metadata. Here, the weights assigned to each item vary depending on the genre to which the content belongs, and in particular, a higher weight is assigned to correspond to the preferences of the majority of users within that genre.

[0043] In step 320, the electronic device (100) identifies content viewed by the user based on the user's viewing history of the electronic device (100). Here, the identified content may be one or more.

[0044] In step 330, the electronic device (100) calculates a recommendation score for each of the multiple contents based on the similarity of the values ​​of each item between the multiple recommendable contents and the identified contents, and the weight assigned to each item by genre. The multiple contents referred to in this step are the contents that the electronic device (100) can obtain.

[0045] Specifically, the electronic device (100) calculates a similarity for each item by comparing the values ​​for each item between a predetermined first content and a second content identified as having been viewed by a user of the electronic device (100) among a plurality of contents, and individually applies a weight corresponding to each item to the similarity for each item, and calculates the total sum as the recommendation score of the first content. The electronic device (100) calculates a recommendation score for each of the plurality of contents in this manner.

[0046] In step 340, the electronic device (100) performs a recommendation operation regarding content with a high recommendation score among multiple contents. Here, one or more contents with a high recommendation score may be selected. For example, the electronic device (100) may display a UI that recommends content with a recommendation score higher than a preset rank, or may display a UI that includes a recommendation menu for the content with the highest recommendation score.

[0047] Thus, the electronic device (100) can more easily identify recommended content that matches the preferences of the user of the electronic device (100) among a plurality of contents.

[0048] Meanwhile, the processor (170) of the electronic device (100) may perform at least some of the data analysis, processing, and result information generation for performing the operation of calculating recommendation scores for each of a plurality of contents using the information as described above, using at least one of machine learning, neural network, or deep learning algorithms as a rule-based or artificial intelligence (AI) algorithm. For example, the information in step 320 may be designed as an AI model, and the electronic device (100) may identify recommended content by inputting information regarding the user's viewing history into the AI ​​model.

[0049] For example, the processor (170) of the electronic device (100) can perform the functions of both a learning unit and a recognition unit. The learning unit performs the function of generating a learned neural network, and the recognition unit performs the function of recognizing (or inferring, predicting, estimating, or judging) data using the learned neural network. The learning unit can generate or update the neural network. The learning unit can acquire learning data to generate the neural network. For example, the learning unit can acquire learning data from the storage unit of the electronic device (100) or server (200) or from an external source. The learning data may be data used for training the neural network, and the neural network can be trained using the data that has performed the above-mentioned operation as learning data.

[0050] The learning unit may perform preprocessing on the acquired training data or select data to be used for training from among multiple training data before training a neural network using the training data. For example, the learning unit may process the training data into a pre-set format, filter it, or add / remove noise to process it into a form suitable for training. The learning unit may generate a neural network configured to perform the above operations using the preprocessed training data.

[0051] A trained neural network may be composed of multiple neural networks (or layers). The nodes of the multiple neural networks have weights, and the multiple neural networks may be connected to each other such that the output value of one neural network is used as the input value of another neural network. Examples of neural networks may include models such as CNN (Convolutional Neural Network), DNN (Deep Neural Network), RNN (Recurrent Neural Network), RBM (Restricted Boltzmann Machine), DBN (Deep Belief Network), BRDNN (Bidirectional Recurrent Deep Neural Network), and Deep Q-Networks.

[0052] Meanwhile, the recognition unit may acquire target data to perform the above-mentioned operation. The target data may be acquired from a storage unit of an electronic device (100) or a server (200), or from an external source. The target data may be data to be recognized by a neural network. Before applying the target data to a trained neural network, the recognition unit may perform preprocessing on the acquired target data or select data to be used for recognition from among a plurality of target data. For example, the recognition unit may process the target data into a pre-set format, filter it, or add / remove noise to process it into a form of data suitable for recognition. By applying the preprocessed target data to the neural network, the recognition unit may acquire an output value output from the neural network. Along with the output value, the recognition unit may acquire a probability value or a confidence value.

[0053] Hereinafter, a specific method for an electronic device (100) to identify recommended content is described.

[0054] Figure 4 is a block diagram showing the schematic operation of an electronic device.

[0055] As illustrated in FIGS. 1, 3 and 4, the operation according to an embodiment of the present invention is divided into a plurality of steps, such as a second information generation unit (410), a second information-based processing unit (420), a first information input unit (430), an output unit (440), and a user tendency inference unit (450). In this embodiment, the first information is information regarding the user's viewing history at step 320 of FIG. 3, and the second information is information regarding the weight at step 310.

[0056] The above-described components are classified according to functional operations and are generally performed by the processor (170) of the electronic device (100). However, depending on the design method, some operations may be performed by the server (200); for example, the operation of the second information generation unit (410) may be performed by the server (200). Among the steps of FIG. 3 above, step 320 relates to the operation of the first information input unit (430), step 310 to the operation of the second information generation unit (410), step 330 to the operation of the second information-based processing unit (420), and step 340 to the operation of the output unit (440). This embodiment describes the case where the operation of the user tendency inference unit (450) is included, but depending on the design method, it is also possible to not include the operation of the user tendency inference unit (450).

[0057] The second information generation unit (410) generates second information based on big data of the viewing history of many users for multiple contents. The second information according to the present embodiment is an AI model or AI algorithm based on the concept of a decision tree, and is configured to identify recommended content corresponding to the first information among multiple contents when a predetermined first information is input. The second information generation unit (410) generates or updates the second information through learning based on the aforementioned big data.

[0058] Here, each of the multiple contents possesses metadata to describe the characteristics of the content. The metadata refers to a specific value for each of the multiple items. The characteristics of the content are expressed by these item-specific values. Additionally, each piece of content belongs to a specific genre, and the genre of the content is indicated by the value of the genre item in the metadata.

[0059] The second information of the present embodiment includes a weight individually assigned to each of a plurality of items for indicating the characteristics of the content, corresponding to the genre to which the content belongs. When the second information is a decision tree-based model, these weights are assigned to each node within the second information, thereby determining the path of data passing between the nodes within the second information. The path of data within the second information determines the identification result of the recommended content.

[0060] The second information generation unit (410) enables the content viewing history of various users corresponding to each of multiple genres to be reflected in the second information through learning of big data. For example, the server (200) obtains the viewing history information of the users of the corresponding clients from a number of connectable clients. Based on the information obtained in this manner, the server (200) assigns a weight to each item of content belonging to a predetermined genre, reflecting the preference characteristics of the users of the multiple clients for that genre. Specific examples of assigning weights to multiple items corresponding to a specific genre will be described later.

[0061] The second information-based processing unit (420) identifies one or more recommended contents among multiple contents by using the second information generated by the second information generation unit (410) and the user's first information transmitted from the first information input unit (430). In the second information, weights are assigned to each of the multiple items of a predetermined content, and these weights are determined according to the genre to which the content belongs. Therefore, if the genre changes, the weight for the same item also changes. The second information-based processing unit (420) calculates the similarity between each first content among the multiple contents and the items of the content obtained from the first information (i.e., the content viewed by the user), calculates the item-specific weights of the first content by reflecting them in the calculated similarity, calculates the item-specific scores, and calculates the recommendation score of the first content by summing all the item-specific scores. The second information-based processing unit (420) calculates the recommendation score for each of the multiple contents according to this method.

[0062] The second information-based processing unit (420) transmits one or more contents with high recommendation scores among multiple contents to the output unit (440) as recommended content.

[0063] The first information input unit (430) transmits first information based on the viewing history of the user of the electronic device (100) to the second information-based processing unit (420). The first information input unit (430) generates the first information in accordance with a predefined format so that it can be input into the second information. Based on the user's viewing history, the first information includes the values ​​of each of a plurality of items regarding one or more contents viewed by the user. Additionally, the first information input unit (430) provides information related to a plurality of contents that can be provided to the electronic device (100) at the current time (e.g., EPG (electronic program guide) information) to the second information-based processing unit (420), thereby enabling the second information-based processing unit (420) to identify recommended content among the plurality of contents based on the information.

[0064] The output unit (440) outputs one or more recommended contents transmitted from the second information-based processing unit (420). For example, the output unit (440) may display a UI (10, see FIG. 2) that guides the identified recommended contents.

[0065] The user preference inference unit (450), separately from the second information-based processing unit (420), identifies content characteristics that are highly preferred by the user based on the first information, without being limited to a specific genre, as an inference result. The inference result may be, for example, content or genres that are highly preferred by the user (i.e., frequently viewed by the user or have a long viewing time). The user preference inference unit (450) transmits the identified inference result to the second information-based processing unit (420) or the output unit (440) so that it acts as a filter for the identified recommended content. For example, when 30 recommended contents are identified, the second information-based processing unit (420) may exclude 10 recommended contents that have the lowest relevance to the content characteristics according to the inference result from the user preference inference unit (450) and transmit the remaining 20 recommended contents to the output unit (440). Depending on the design method, the user preference inference unit (450) may be excluded from the entire component.

[0066] The metadata of the content is explained below.

[0067] Figure 5 is an example diagram showing the data structure of each content.

[0068] As illustrated in FIGS. 1 and 5, a plurality of contents (510, 520, 530) obtainable by the electronic device (100) each include main data (511) and metadata (512). In this embodiment, the case of a predetermined first content (510) is described, but the remaining contents (520, 530) are also similar to the case of the first content (510). The main data (511) represents the first content (510) itself, and corresponds to video data when the first content (510) is a video, photo data when the first content (510) is a photo, and game content data when the first content (510) is a game.

[0069] Metadata (512) includes various additional information regarding the main data (511), that is, information describing the main data (511). Metadata (512) indicates the characteristics of the first content (510) by including the values ​​of the first content (510) for each of the multiple predefined items. In this embodiment, examples of some of the multiple items are described, but these examples do not limit the types of items. For example, Title indicates the title of the first content (510), Sub_genre indicates the detailed genre of the first content (510), Channel_number indicates the channel number where the first content (510) is broadcast, Program_id indicates the ID assigned to the first content (510), Series_id indicates the ID of the series if the first content (510) is included in a series, Description indicates a description regarding the first content (510), etc. Metadata (512) may also include values ​​for various other types of items. Metadata (512) containing the values ​​of each of these items represents the characteristics of the first content (510). The remaining content (520, 530) can also represent the characteristics of the content (520, 530) in this manner.

[0070] As such, the characteristics of each content (510, 520, 530) are represented by multiple values ​​corresponding to multiple items, and the preference for each item (specifically, the preference of multiple users who have watched the content (510, 520, 530), based on big data of the viewing history of various users) differs in accordance with the genre to which the content (510, 520, 530) belongs. The following describes such an embodiment.

[0071] Figure 6 is a table showing the characteristics of content genre preferences and the items corresponding to each characteristic.

[0072] As illustrated in Fig. 6, the genres of content are diverse, including sports, news, movies, dramas, entertainment, animation, and music. Additionally, a single piece of content belongs to a single genre. A single piece of content may belong to multiple genres, in which case it is considered to belong to one representative genre among the multiple genres. Here, according to big data based on the viewing history of various users, the characteristics of content highly preferred by various users differ depending on the genre of the content.

[0073] For example, in the sports genre, characteristics highly preferred by users include specific sports (baseball, soccer, basketball, etc.), professional leagues of those sports (in the case of soccer, professional soccer leagues in various countries such as Korea, the UK, Germany, France, Spain, and Italy), and teams in those sports. In the news genre, characteristics highly preferred by users include specific news programs and the channels on which those news programs are broadcast. In the drama genre, characteristics highly preferred by users include series and genre preferences.

[0074] Additionally, each attribute corresponds to one or more items among multiple items in the metadata. In this case, a single attribute does not necessarily correspond to only one item; multiple attributes may correspond to a single item, or a single attribute may correspond to multiple items. For example, the series attribute corresponds to the series_id item. The channel attribute corresponds to the channel_number item. The program attribute corresponds to the title and program_id attributes.

[0075] Consequently, for each genre, there exists one or more items that are relatively highly preferred by multiple users. An item corresponding to a specific genre is an item associated with content that is highly preferred within that genre. For example, the items corresponding to the highly preferred traits in the sports genre are sub_genre and title. The items corresponding to the highly preferred traits in the drama genre are series_id and sub_genre. The items corresponding to the highly preferred traits in the animation genre are series_id, channel_number, title, and description.

[0076] In this way, items corresponding to characteristics that are highly preferred in a specific genre are assigned weights differently from items that are not. For example, when the genre of a piece of content is sports, the weight assigned to sub_genre and title among multiple items is relatively higher than the weight assigned to the remaining items. In the second information described above, weights are assigned in this manner for all content that can be obtained by the electronic device (100). Below, an example of assigning weights to each genre in the second information will be described.

[0077] Figure 7 is a table showing examples of weights assigned to each item by genre.

[0078] As illustrated in FIGS. 6 and 7, when there are multiple items sub_genre, program_id, series_id, title, channel_number, and description, the weights assigned to each item differ by genre. When there are multiple items corresponding to a single genre (i.e., highly preferred in a single genre), there may be differences in weights among these items depending on their preference. Weights may be assigned only to items corresponding to a single genre, or weights may be assigned to the remaining items as well. In the case where weights are assigned to the remaining items, the weight for highly preferred items is set higher than the weight for items that are not highly preferred.

[0079] For example, in the case of the sports genre, weights of 0.5 and 0.2 are assigned to the items sub_genre and title, respectively, which correspond to the highly preferred trait. In other words, it can be seen that in the sports genre, the trait corresponding to sub_genre has a higher preference than the trait corresponding to title. Additionally, among the remaining items, a weight of 0.1 is assigned to channel_number, which is related to the trait with the next highest preference after sub_genre and title. Here, the numerical values ​​of the weights can be varied and determined depending on the design method and are merely an example. In the case of the entertainment genre, equal weights of 0.5 are assigned to the items series_id and title, respectively, which correspond to the highly preferred trait. That is, in the entertainment genre, the preference of the trait corresponding to series_id is the same as the preference of the trait corresponding to title. In the case of the animation genre, differential weights of 0.4, 0.2, 0.2, and 0.1 are assigned to the items series_id, title, channel_number, and description, respectively, which correspond to the highly preferred trait.

[0080] The method of assigning weights is not limited; for example, the total sum of weights assigned to all items can be set to 1, and the weight of each item can be set differently depending on the item's preference. For example, in the case of the entertainment genre, a weight of 0.5 is assigned to each of series_id and title, and no weight is assigned to the remaining items.

[0081] For items in the table of this drawing for which no weight is specified, a weight lower than that of an item for which a weight is specified may be applied, or no weight may be applied. For example, in the case of a sports genre, a weight of 0 may be applied to program_id, series_id, and description.

[0082] Figure 8 is a table showing an example of a user's viewing history of an electronic device.

[0083] As illustrated in FIG. 8, the first information is generated based on the viewing history of the user of the electronic device (100) accumulated over a predetermined period of time. From the metadata of multiple contents viewed by the user, values ​​corresponding to each item of the metadata may be represented in a table format accumulated for each content. Since the user's viewing history includes the history of one or more contents viewed by the user, the first information may be the accumulation of metadata of multiple contents viewed by the user. Accordingly, the first information includes one or more values ​​corresponding to each of a plurality of items.

[0084] The viewing history table of the present drawing includes details such as the viewing start time, viewing end time, title (corresponding to item title), genre (corresponding to item sub_genre), and content ID (corresponding to item program_id) of the content viewed by the user. One row of the table corresponds to the history of viewing a specific piece of content once. One column of the table represents the value of each of multiple pieces of content corresponding to one item of metadata. For example, according to the first row of the table, it indicates that on August 26, 2020, for about 30 minutes, the user watched content with the title "OOOO Cup Playoff," the genre "Hockey," and the content ID "ba000000000003933376".

[0085] Based on the metadata of various content viewed by the user, a table may be created having multiple values ​​corresponding to each item of the metadata (since multiple contents will each have unique values ​​for the same item). However, the data does not necessarily have to be stored in a table form in the actual electronic device (100). The first information may be prepared in various ways based on data such as such a table.

[0086] For example, the electronic device (100) may use metadata of one or more contents having a viewing frequency higher than a threshold value (one content with the highest viewing frequency, or multiple contents having a viewing frequency higher than a predetermined rank) as first information.

[0087] Alternatively, the electronic device (100) may select one or more of the content among multiple contents through genre filtering and use the metadata of the selected one or more contents as first information. For example, the electronic device (100) may select the content of the genre with the highest viewing frequency among the various contents included in the viewing history.

[0088] When the first information is prepared in this manner, the electronic device (100) identifies the degree of similarity between each item by comparing each of the multiple obtainable contents with the first information. An embodiment thereof is described below.

[0089] Figure 9 is an example diagram showing a method for calculating item-by-item similarity between recommendable content and viewed content.

[0090] As illustrated in FIG. 1 and FIG. 9, the electronic device (100) can calculate the item-specific similarity (930) between the metadata (910) of a predetermined first content among a plurality of recommendable content acquired by the electronic device (100) and the metadata (920) of a predetermined second content viewed by the user represented by the first information. Since the metadata (910) of the first content includes a plurality of item-specific values ​​and similarly the metadata (920) of the second content also includes a plurality of item-specific values, the electronic device (100) can calculate the similarity (930) between these two values ​​by comparing two values ​​of the same item.

[0091] For example, when there are values ​​of a first content and a second content for each of multiple items #A, #B, #C, #D, #E, and #F, the electronic device (100) compares the values ​​of the first content and the second content in the same item. The electronic device (100) compares the value of the first content S1 and the value of the second content V1 in item #A to calculate the similarity L1 between S1 and V1 in item #A. The electronic device (100) compares the value of the first content S2 and the value of the second content V2 in item #B to calculate the similarity L2 between S2 and V2 in item #B. The electronic device (100) calculates the similarity for the remaining items according to the same principle. As similarity is calculated for each item in this way, the similarity (930) between the first content and the second content is finally derived.

[0092] The similarity between two values ​​for each item can be calculated according to various principles and methods. For example, if the item values ​​S1 and V1 are numbers, the smaller the difference between the two values ​​S1 and V1, the higher the similarity L1; conversely, the larger the difference, the lower the similarity L1. If the item values ​​S1 and V1 are text or binary code, the more common keywords or codes there are between the two values ​​S1 and V1, the higher the similarity L1; conversely, the fewer common keywords or codes there are, the lower the similarity L1. If the two values ​​S1 and V1 can be represented as coordinates or vectors on a Euclidean plane, the smaller the Euclidean distance between the two coordinates or vectors corresponding to S1 and V1, respectively, the higher the similarity L1; conversely, the longer the Euclidean distance, the lower the similarity L1. As such, any mathematical method can be applied without limitation, as long as the similarity between two values ​​for each item can be represented by quantitative numerical values.

[0093] Figure 10 is an example diagram showing a method for calculating a recommendation score for the content by reflecting weights in the calculated similarity.

[0094] As illustrated in FIGS. 1 and FIGS. 10, the electronic device (100) calculates a similarity (930) between a first content and a second content, and applies an item-specific weight (940) corresponding to the genre of the first content, which is the subject of the recommendation score calculation, to the similarity (930). Such weights (940) are provided in the second information. That is, the electronic device (100) calculates a recommendation score for a content based on the first information and the second information.

[0095] The weight (940) is assigned in advance to each of the multiple items as described in the preceding embodiment. The weight (940) has a unique value corresponding to a specific genre, reflecting the various user preferences for each genre, and the value changes when the genre changes. The weight (940) assigned to each item is reflected in the similarity (930) of the corresponding item.

[0096] For example, the item-specific similarity (930) between the first content and the second content is L1 in item #A, L2 in item #B, L3 in item #C, L4 in item #D, L5 in item #E, and L6 in item #F. The weight (940) corresponding to the genre to which the content belongs is W1 in item #A, W2 in item #B, W3 in item #C, W4 in item #D, W5 in item #E, and W6 in item #F. In this case, the recommendation score Cs of the content can be calculated as follows.

[0097] Cs=(L1*W1)+(L2*W2)+(L3*W3)+(L4*W4)+(L5*W5)+(L6*W6)

[0098] That is, the recommendation score can be expressed as the sum of the products of the degrees of freedom (930) of each item and the weight (940) of the corresponding item. If the number of items is N, the recommendation score Cs can be expressed as follows.

[0099] Cs=Σ(L k *W k ); (k is an integer from 1 to N)

[0100] In accordance with this method, the electronic device (100) can calculate a recommendation score for each of the multiple available contents.

[0101] The method for calculating such recommendation scores may vary depending on the form of the item-specific values ​​included in the first information. The example described above applies when there is only one item-specific value in the first information (when the first information contains only one piece of content); however, alternatively, there may be multiple item-specific values ​​in the first information (when the first information contains multiple pieces of content). In this case, the multiple item-specific values ​​of the first information are individually compared with the values ​​of the same items in the first content being compared to calculate multiple similarities, and weights are applied to each of the multiple similarities to sum them up.

[0102] For example, when the first information for item #A has two values, S11 and S12, and the value of the first content is V1, the electronic device (100) calculates the similarity L11 between S11 and V1 and the similarity L12 between S12 and V1. The electronic device (100) applies a genre-specific weight to each similarity for item #A, such as (L11+L12)*W1. Additionally, when the first information for item #B has two values, S21 and S22, and the value of the first content is V2, the electronic device (100) calculates the similarity L21 between S21 and V2 and the similarity L22 between S22 and V2. The electronic device (100) applies a genre-specific weight to each similarity for item #B, such as (L21+L22)*W2. In this way, the recommendation score Cs can be obtained by summing the results reflecting weights for all items (e.g., (L11+L12)*W1+(L21+L22)*W2+...).

[0103] When the electronic device (100) obtains a recommendation score for each content in this manner, it identifies recommended content among multiple contents based on each obtained recommendation score. The method for identifying recommended content is as described above.

[0104] Meanwhile, the electronic device (100) may identify content belonging to the same genre as the content viewed by the user in the first information among multiple recommendable contents, and calculate a recommendation score for each content by comparing the identified content with the content viewed by the user in the first information. For example, if the viewing history included in the first information is of the sports genre, the electronic device (100) first identifies content belonging to the sports genre among multiple recommendable contents and obtains a recommendation score for each of the identified contents.

[0105] Alternatively, the electronic device (100) may provide an environment for selecting metadata of a specific genre from the first information when the first information includes metadata of content of various genres. The following describes such an embodiment.

[0106] Figure 11 is an example of an electronic device displaying a UI for genre selection.

[0107] As illustrated in FIG. 11, the electronic device (100) displays a UI (1100) that allows a user to select one of a plurality of genres. The UI (1100) provides an option for selecting one of a plurality of genres, such as sports, news, movies, dramas, entertainment, animation, and music.

[0108] When a genre is selected in the UI (1100), the electronic device (100) obtains first information regarding the content of the selected genre from the user's viewing history. Based on the first information obtained in this way, the electronic device (100) calculates a recommendation score for each of the multiple available contents. Based on the recommendation scores of each content calculated, the electronic device (100) identifies one or more recommended contents and displays a UI (1100) that guides the identified recommended contents.

[0109] Methods according to exemplary embodiments of the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. Such a computer-readable medium may include program instructions, data files, data structures, etc., either alone or in combination. For example, the computer-readable medium may be stored in a non-volatile storage device such as a USB memory device, regardless of whether it is erasable or rewritable, or in a memory such as, for example, RAM, ROM, flash memory, memory chip, integrated circuit, or a storage medium that is optically or magnetically writable and simultaneously machine-readable (e.g., computer), such as, for example, CD, DVD, magnetic disk, or magnetic tape. It will be understood that memory included in a mobile terminal is an example of a machine-readable storage medium suitable for storing a program or programs containing instructions for implementing embodiments of the present invention. The program instructions recorded on the storage medium may be those specifically designed and configured for the present invention, or may be those known and available to a person skilled in the art of computer software. Alternatively, the computer program instructions may be implemented by a computer program product. Explanation of the symbols

[0110] 100 : Electronic device 120 : Display section 170 : Processor

Claims

Claim 1 An electronic device comprises a processor that obtains information on weights assigned to each of a plurality of items having values ​​representing content characteristics, wherein the weights are assigned highly to items corresponding to content characteristics preferred by a plurality of users according to the genre of content among the plurality of items, identifies content viewed by the user based on the viewing history of the user of the electronic device, calculates a recommendation score for each of the plurality of content based on the similarity of values ​​of each item between a plurality of recommendable content and the identified content, and the weights assigned to each item according to the genre, and performs a recommendation operation regarding the content with a high calculated recommendation score among the plurality of recommendable content, wherein the processor calculates a similarity by item by comparing the values ​​of each of the plurality of items between a predetermined first content among the plurality of content and a second content identified as having been viewed by the user, identifies the weights corresponding to each item, and calculates a total value by reflecting each identified weight in the similarity of each item as the recommendation score of the first content. Claim 2 delete Claim 3 An electronic device according to claim 1, wherein the value of the first item among the plurality of items is represented as a numerical value, and the processor identifies a higher degree of similarity for each first item to the first content as the difference between the first content and the second content is smaller. Claim 4 An electronic device according to claim 1, wherein the value of the second item among the plurality of items appears as text, and the processor identifies a higher degree of similarity for each second item to the first content as the more common keywords there are between the values ​​of the second item of each of the first content and the second content. Claim 5 In claim 1, the processor is an electronic device that identifies the content with the highest viewing frequency among the content included in the viewing history as the second content. Claim 6 In claim 1, the processor is an electronic device that identifies the genre with the highest viewing frequency among the content included in the viewing history and identifies the content of the identified genre as the second content. Claim 7 In claim 1, the processor is an electronic device that displays a UI on a display that enables the selection of one or more contents having a recommendation score higher than a preset rank among the plurality of contents. Claim 8 In claim 1, the processor is an electronic device that identifies a plurality of recommendable contents based on EPG (electronic program guide) information. Claim 9 A method for controlling an electronic device comprises: a step of obtaining information on weights assigned to each of a plurality of items having values ​​representing content characteristics; wherein the weights are assigned highly to items among the plurality of items that correspond to content characteristics preferred by a plurality of users according to the genre of content; a step of identifying content viewed by a user based on the viewing history of a user of the electronic device; a step of calculating a recommendation score for each of the plurality of contents based on the similarity of the values ​​of each item between a plurality of recommendable contents and the identified content, and the weights assigned to each item according to the genre; and a step of performing a recommendation operation for content with a high recommendation score among the plurality of recommendable contents. The step of calculating the recommendation score for each of the plurality of contents comprises: a step of calculating item-specific similarity by mutually comparing the values ​​of each of the plurality of items between a predetermined first content among the plurality of contents and a second content identified as having been viewed by the user; a step of identifying the weights corresponding to each item; and a step of calculating the recommendation score of the first content by reflecting each identified weight in the item-specific similarity and summing the values. A method for controlling an electronic device including Claim 10 delete Claim 11 A control method for an electronic device according to claim 9, wherein the value of the first item among the plurality of items is represented as a numerical value, and the step of calculating the similarity includes the step of identifying the similarity for each first item for the first content as high as the difference between the value of the first item between the first content and the second content is small. Claim 12 A control method for an electronic device according to claim 9, wherein the value of the second item among the plurality of items is displayed as text, and the step of calculating the similarity includes the step of identifying the similarity for each second item to the first content as high as there are more common keywords between the values ​​of the second item of each of the first content and the second content. Claim 13 ◈Claim 13 was abandoned upon payment of the registration fee.◈ In claim 9, the step of identifying the content viewed by the user comprises the step of identifying the content with the highest viewing frequency among the content included in the viewing history as the second content. A method for controlling an electronic device. Claim 14 ◈Claim 14 was abandoned upon payment of the registration fee.◈ In claim 9, the step of identifying the content viewed by the user comprises identifying the genre with the highest viewing frequency among the content included in the viewing history, and identifying the content of the identified genre as the second content. A method for controlling an electronic device. Claim 15 ◈Claim 15 was abandoned upon payment of the registration fee.◈ In claim 9, the step of performing the recommendation operation comprises a method for controlling an electronic device that includes the step of displaying on a display a UI that enables the selection of one or more contents having a recommendation score of a rank higher than or equal to a preset rank among the plurality of contents.

Images