Live broadcast method and device based on digital human, electronic equipment and storage medium

Abstract

The invention provides a live broadcast method and device based on digital people, electronic equipment and a storage medium, and relates to the field of artificial intelligence, in particular to the field of digital people and the field of live broadcast. According to the specific implementation scheme, the method comprises the steps of determining to-be-broadcasted content from a preset content database; wherein the content to be broadcasted represents content information of an event to be broadcasted; extracting live broadcast information from the content to be broadcasted; wherein the live broadcast information represents key information in the content to be broadcasted; and playing the live broadcast information in a pre-established live broadcast room based on a preset image of the digital human. By automatically determining the content of the event to be broadcasted and automatically converting the content into the live broadcast information to be displayed in the live broadcast room, the content is automatically broadcasted through the digital person, and the watching experience of a user is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of digital humans and live broadcasting in the field of artificial intelligence, and in particular to a live broadcasting method, device, electronic device and storage medium based on digital humans. Background Art

[0002] Users can view hot content at any time through tools such as social networking sites or video platforms. Hot content is generally displayed in the form of videos or pictures and texts, and users can view it by clicking or sliding.

[0003] The current content display format is not vivid enough, and users are not very interested in watching it. Therefore, it is necessary to adopt a more flexible and interesting way to display the content to improve the user experience. Summary of the invention

[0004] The present disclosure provides a live broadcast method, device, electronic device and storage medium based on digital human.

[0005] According to a first aspect of the present disclosure, a live broadcast method based on a digital human is provided, comprising:

[0006] Determine the content to be broadcast from a preset content database; wherein the content to be broadcast represents content information of the event to be broadcast;

[0007] Extracting live broadcast information from the content to be broadcast; wherein the live broadcast information represents key information in the content to be broadcast;

[0008] Based on the preset digital human image, the live broadcast information is played in a pre-built live broadcast room.

[0009] According to a second aspect of the present disclosure, a live broadcast device based on a digital human is provided, comprising:

[0010] A content determination unit, configured to determine the content to be broadcast from a preset content database; wherein the content to be broadcast represents content information of the event to be broadcast;

[0011] An information extraction unit, configured to extract live broadcast information from the content to be broadcast; wherein the live broadcast information represents key information in the content to be broadcast;

[0012] The live broadcast unit is used to play the live broadcast information in a pre-built live broadcast room based on the preset digital human image.

[0013] According to a third aspect of the present disclosure, there is provided an electronic device, including:

[0014] at least one processor; and

[0015] a memory communicatively coupled to the at least one processor;

[0016] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect of the present disclosure.

[0017] According to a fourth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause the computer to execute the method according to the first aspect of the present disclosure.

[0018] According to a fifth aspect of the present disclosure, a computer program product is provided, comprising a computer program, which implements the steps of the method described in the first aspect of the present disclosure when executed by a processor.

[0019] According to the technology disclosed in the present invention, it is possible to automatically broadcast content through digital humans, thereby improving the viewing experience of users.

[0020] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0022] Figure 1 It is a flowchart of a live broadcast method based on digital human provided according to an embodiment of the present disclosure;

[0023] Figure 2 is a schematic diagram of a live broadcast room provided according to an embodiment of the present disclosure;

[0024] Figure 3 It is a flowchart of a live broadcast method based on digital human provided according to an embodiment of the present disclosure;

[0025] Figure 4 It is a flowchart of a live broadcast method based on digital human provided according to an embodiment of the present disclosure;

[0026] Figure 5 is a schematic diagram of a live broadcast room provided according to an embodiment of the present disclosure;

[0027] Figure 6 It is a structural block diagram of a live broadcast device based on a digital human provided according to an embodiment of the present disclosure;

[0028] Figure 7 It is a structural block diagram of a live broadcast device based on a digital human provided according to an embodiment of the present disclosure;

[0029] Figure 8 is a block diagram of an electronic device for implementing the live broadcast method based on digital human according to an embodiment of the present disclosure;

[0030] Fig. 9 It is a block diagram of an electronic device used to implement the live broadcast method based on digital human in the embodiment of the present disclosure. DETAILED DESCRIPTION

[0031] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0032] Currently, hot content is generally displayed in the form of pictures, texts or videos, and users need to click or slide the hot content one by one to view it. For example, a user can click the title of a hot content to view the content. The presentation of hot content is not vivid enough, and users are not very interested in watching it.

[0033] In addition, users can post comments or opinions on the content they are viewing. Currently, most users communicate by posting comments under hot content, which has a certain delay. For example, if a user posts a comment through his own terminal device, the comment cannot appear directly on another user's terminal device, resulting in the inability of users to interact in real time on hot content, and the user experience is poor.

[0034] Currently, there is a lack of live broadcasting for hot content in the market. If live broadcasting is used, each hot content needs to have a corresponding copywriting script, and the writing of the copywriting will also waste a lot of manpower and time, making the live broadcast effect less than ideal.

[0035] The present disclosure provides a digital human-based live broadcast method, device, electronic device and storage medium, which are applied to the digital human field and the live broadcast field in the field of artificial intelligence to realize automatic broadcasting of content through digital humans and enhance the user's viewing experience.

[0036] It should be noted that the model in this embodiment is not a model for a specific user and cannot reflect the personal information of a specific user. It should be noted that the data in this embodiment comes from a public data set.

[0037] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0038] In order to enable readers to more deeply understand the implementation principle of the present disclosure, the following Figure 1-Figure 9 The embodiment is further refined.

[0039] Figure 1 FIG. 1 is a flow chart of a digital human-based live broadcast method according to an embodiment of the present disclosure, which can be performed by a digital human-based live broadcast device. Figure 1 As shown, the method comprises the following steps:

[0040] S101. Determine content to be broadcast from a preset content database; wherein the content to be broadcast represents content information of an event to be broadcast.

[0041] For example, different events may occur every day, and each event has its own content information, which may refer to descriptions and comments on the event. Users can obtain content information of various events through media channels such as websites and social platforms. A content database is pre-set, and the content database may store content information of events within a period of time. For example, the content information of events occurring within 24 hours may be stored, so that the events in the content database are current hot events.

[0042] The content database can store content information of multiple events. When broadcasting live, one or more events can be selected and reported. The content to be reported can be automatically determined from the preset content database. The content to be reported refers to the content information of the event to be reported. There can be one or more events to be reported, that is, multiple contents to be reported can be determined. For example, multiple contents to be reported can be randomly determined from all the contents in the preset content database, or the content with the highest number of views can be determined as the content to be reported based on the number of views of each content in the content database within one minute after being released.

[0043] The content information in the content database may be in the form of pictures, texts, videos, etc., and is used to represent the relevant circumstances of the event. For example, the content information may be photos of the scene of the event, reports on the event, comments on the event by netizens, etc. That is, the content to be reported may represent the circumstances of the event to be reported.

[0044] S102: extracting live broadcast information from the content to be broadcast; wherein the live broadcast information represents key information in the content to be broadcast.

[0045] Exemplarily, the content to be broadcast may be in various forms, and there may be redundant information in the content to be broadcast. When broadcasting live, if the content to be broadcast is played or displayed directly, the live broadcast effect may be affected. Therefore, it is necessary to process the content to be broadcast and extract live broadcast information suitable for live broadcast from the content to be broadcast. The live broadcast information can represent key information in the content to be broadcast. For example, keywords can be extracted from the content to be broadcast, and then the keywords can be reorganized into text data as live broadcast information.

[0046] The content to be broadcasted may also be semantically processed to obtain an explanation of the content to be broadcasted, and a corresponding image may be generated based on the explanation, and the image may be used as live broadcast information. In this embodiment, the live broadcast information may include a title, a text, a background image, etc., that is, based on the content to be broadcasted, a title and a script text may be obtained when the event to be broadcasted is broadcasted, and a background image may be obtained to be displayed to the user.

[0047] S103: Based on the preset digital human image, broadcast live information in a pre-built live broadcast room.

[0048] For example, the digital human can be a virtual character image or a real character image. The image of the digital human is preset and the digital human is used as the anchor of the live broadcast room. A live broadcast room is pre-built, and the live broadcast room can be a real scene or a virtual scene, and the digital human is displayed in the live broadcast room.

[0049] For each event to be broadcast, the live broadcast information can be displayed or played in the live broadcast room. For example, if the live broadcast information is the script text of the event to be broadcast, the script text can be converted into voice data, and the digital human plays the voice data so that the user can hear the script text. In this embodiment, the text-to-speech process is not specifically limited. When the digital human plays the voice data, the mouth shape and body shape can also change accordingly, making the live broadcast more vivid.

[0050] If the live broadcast information is a background image, the background image can be displayed as the background of the live broadcast room. For example, if the event to be broadcast is about a certain animal, the background of the live broadcast room can be a photo of the animal. If the live broadcast information is a title, the title can be superimposed on the background image so that the user can clearly understand the event currently being broadcast.

[0051] In this embodiment, live broadcast information is played in a pre-built live broadcast room, including: if multiple contents to be broadcast are determined from a preset content database, the live broadcast information corresponding to the contents to be broadcast are played in sequence in the pre-built live broadcast room according to the storage order of the contents to be broadcast in the preset content database.

[0052] Specifically, one or more contents to be broadcasted can be determined from a preset content database. If one content to be broadcasted is determined, after the live broadcast information of the content to be broadcasted is determined, the live broadcast information is played, and the live broadcast ends after the playing is completed.

[0053] If multiple contents to be broadcast are determined, the live broadcast information of each content to be broadcast can be determined. When playing the live broadcast information, each live broadcast information can be played in sequence, that is, after playing the live broadcast information of one content to be broadcast, the live broadcast information of the next content to be broadcast is played directly until the live broadcast information of all contents to be broadcast is played and the live broadcast ends.

[0054] The live broadcast order of each content to be broadcast can be preset. For example, the storage order of the content to be broadcast in the preset content database can be used as the live broadcast order. That is, the live broadcast information corresponding to the content to be broadcast is played in sequence according to the storage order of the content to be broadcast in the preset content database. Each time the live broadcast information of a content to be broadcast is played, the background image and title displayed in the live broadcast room can be changed, and the digital human also plays the corresponding script copy.

[0055] The beneficial effect of such a setting is that the live broadcast information of multiple events can be automatically played in sequence without the need for manual playback control, thus avoiding live broadcast freezes, improving live broadcast efficiency, ensuring live broadcast effects, and enhancing users' viewing experience.

[0056] In this embodiment, it also includes: obtaining question information sent by the user in the live broadcast room; performing semantic recognition processing on the question information to obtain reply information corresponding to the question information; and playing the reply information in a pre-built live broadcast room based on the preset image of the digital human.

[0057] Specifically, when watching the live broadcast, users can also post comments and opinions at any time. For example, if the digital human is explaining a certain animal, the user can ask questions about the specific situation of the animal, that is, send out question information. An input box can be set on the page of the live broadcast room, and the user can send out question information through the input box.

[0058] The question information of each user can be obtained in real time or at a fixed time, and semantic recognition processing can be performed on the question information. For example, the question information can be input into a preset large language model, and the large language model outputs the reply information corresponding to the question information, that is, the answer corresponding to the question information is obtained. In this embodiment, the model structure of the preset large language model is not specifically limited. The digital human can play the reply information to answer the user's questions. For example, after the digital human finishes speaking the live broadcast information of the event, it can answer the question information of the event in sequence, or it can pause the live broadcast information at any time during the broadcast of the live broadcast information, and play the reply information, and continue to play the live broadcast information after the reply information is played.

[0059] After capturing the question information, you can also select important or repeated question information for reply. For example, you can determine the similarity between each question information. If the similarity between at least two question information exceeds the preset similarity threshold, then these at least two question information are regarded as the same question information, and the reply information is determined for reply; if there is a question information that is not similar to any other question information, then this question information will not be replied. Avoid the live broadcast time being too long due to continuous reply, which affects the user's viewing experience.

[0060] The beneficial effect of this arrangement is that by capturing question information and answering it, real-time communication can be carried out with users, and different users can also communicate in real time through comments, which increases the fun of the broadcast and enhances the user experience.

[0061] Figure 2 This is a schematic diagram of the live broadcast room. Figure 2 The title of the current live event is displayed in the middle. Users can ask questions through the input box, and all questions asked by users can be displayed in real time in the area above the input box. On the right side of the screen is a digital human who can broadcast live information. A background image can also be displayed after the digital human and the title.

[0062] In the disclosed embodiment, the content information of the event to be broadcast, i.e., the content to be broadcast, is automatically determined from the preset content database, and there is no need to manually confirm what event needs to be broadcast, which reduces manual operations and improves the efficiency of live broadcasting. The content to be broadcasted is automatically processed into live broadcast information that can be displayed in the live broadcast room, enriching the form of data played in the live broadcast room, further saving manpower and time, and improving the fun of the broadcast. Live broadcasting through the preset digital human image makes the live broadcast more vivid and enhances the user's viewing experience.

[0063] Figure 3 A flowchart of a digital human-based live broadcast method provided in an embodiment of the present disclosure.

[0064] In this embodiment, the preset content database includes current hot content, and the current hot content represents content information of the current hot event; determining the content to be broadcast from the preset content database includes: obtaining the browsing information of the current hot content in the preset content database; determining the content to be broadcast from the preset content database according to the browsing information of each current hot content in the preset content database.

[0065] like Figure 3 As shown, the method comprises the following steps:

[0066] S301: Obtain the pageview information of the current hot content in a preset content database.

[0067] Exemplarily, the content in the content database may be current hot content, which represents the content information of current hot events. Current hot events are events that have been discussed recently, for example, events that broke out within 24 hours. Current hot content may be news, reports, videos, voices, discussions, etc. related to the current hot events. Hot events may correspond to different moments, and hot events may be events that have been discussed by netizens, or a threshold of the number of people discussing the event may be preset. If the number of people discussing the event exceeds the threshold of the number of people discussing the event, the event is a hot event.

[0068] Each current hot content has its own pageview information, which indicates the number of times the current hot content has been viewed. For example, it can be the number of times the current hot content has been viewed since it was published, or it can be the number of times the current hot content has been viewed within one hour after it was published. Determine the pageview information of each current hot content in the content database.

[0069] S302: Determine the content to be broadcast from the preset content database according to the pageview information of each current hot content in the preset content database.

[0070] For example, one or more current hot content can be determined from a preset content database as the content to be broadcast according to the pageview information of each current hot content. For example, the current hot content with the highest pageview volume can be determined as the content to be broadcast. Alternatively, the top N current hot content with the highest pageview volume can all be determined as the content to be broadcast.

[0071] In this embodiment, the content to be broadcast is determined based on the number of views, and the events that are currently being discussed hotly can be determined, so that the hot events that users are more concerned about can be broadcast to users, avoiding the broadcast of useless information, reducing the process of users manually searching for hot events, improving the live broadcast effect, and enhancing the user experience.

[0072] In this embodiment, based on the pageview information of each current hot content in the preset content database, the content to be broadcast is determined from the preset content database, including: determining the content level of the current hot content based on the pageview information of the current hot content in the preset content database; wherein the content level represents the degree of influence of the content; and determining the current hot content that is below the preset content level as the content to be broadcast.

[0073] Specifically, the current hot content is graded according to the browsing volume information of the current hot content, and each current hot content corresponds to a content grade. The content grade can represent the impact of the content, and the higher the content grade, the higher the impact. The impact can be the impact on the city, life, work, etc.

[0074] The association between different content levels and pageview information can be preset, and the content level corresponding to the pageview information of the current hot content can be determined according to the preset association, as the content level of the current hot content. For example, the content level can include S level, A level, B level, C level, etc., with S level being the highest and C level being the lowest. S level represents a major event, and the pageviews need to be more than 10 million; A level represents a major event, and the pageviews need to be between 1 million and 10 million; B level represents a general event, and the pageviews need to be between 100,000 and 1 million; C level represents an ordinary event, and the pageviews are less than 100,000.

[0075] A certain content level is pre-determined as a preset content level, and it is determined which current hot content in the content database belongs to the preset content level. The current hot content belonging to the preset content level is determined as the content to be broadcast. For example, the current hot content belonging to the S level is determined as the content to be broadcast.

[0076] The beneficial effect of such a setting is that events are graded according to the number of views, and the content of the highest-level events is used as the content to be broadcast, so as to automatically filter out the content that needs to be broadcast live, reduce the manual screening process, and improve the efficiency of live broadcasting.

[0077] In this embodiment, the current hot content at a preset content level is determined as the content to be broadcast, including: if the number of current hot content at the preset content level is less than a preset quantity threshold, a preset historical database is obtained; wherein the preset historical database includes historical hot content, and the historical hot content represents the content information of hot events occurring within a preset historical time period; according to the quantity difference between the number of current hot content at the preset content level and the preset quantity threshold, the target historical content is determined from the historical hot content in the preset historical database; the current hot content and the target historical content at the preset content level are determined as the content to be broadcast.

[0078] Specifically, the number of current hot content belonging to a preset content level is determined, a number threshold is preset, and the determined number is compared with the preset number threshold. If the number of current hot content belonging to the preset content level is less than the preset number threshold, a preset historical database is obtained. The historical database includes historical hot content, and the historical hot content represents the content information of hot events that occurred within a preset historical time period. For example, the historical hot content can be the content information of hot events in the past month, that is, the content information of hot events that occurred in the past month is stored in the historical database.

[0079] Determine the quantity difference between the quantity of current hot content at a preset content level and a preset quantity threshold, and determine the target historical content from the historical hot content in a preset historical database according to the quantity difference. For example, if the quantity difference is 2, two historical hot content can be obtained from the historical database as the target historical content. In this embodiment, the historical hot content can be obtained randomly or in the order of the historical hot content. The current hot content at the preset content level and the obtained target historical content are both determined as the content to be broadcast.

[0080] If the number of current hot content belonging to the preset content level is equal to or greater than the preset number threshold, there is no need to obtain the historical database, and the current hot content belonging to the preset content level is determined as the content to be broadcast.

[0081] The beneficial effect of such a setting is that if the amount of content to be broadcast determined from the content database is insufficient, it can be supplemented from historical hot content to avoid the live broadcast time being too short and affecting the user's viewing experience.

[0082] In this embodiment, target historical content is determined from historical hot content in a preset historical database based on the quantity difference between the quantity of current hot content at a preset content level and a preset quantity threshold, including: obtaining the pageview information of the historical hot content in the preset historical database; sorting the historical hot content based on the pageview information of each historical hot content in the preset historical database; determining the target historical content based on the sorting result and the quantity difference between the quantity of current hot content at a preset content level and a preset quantity threshold.

[0083] Specifically, the history database includes multiple historical hot content, each of which has its own pageview information, and the pageview information of each historical hot content in the history database is determined. The historical hot content is sorted according to the pageview information of the historical hot content. For example, the pageviews can be sorted in descending order.

[0084] Determine the quantity difference between the quantity of current hot content at a preset content level and a preset quantity threshold, and determine the target historical content from the historical database according to the sorting result and the quantity difference. According to the sorting result, starting from the historical hot content ranked first, the historical hot content with the quantity difference can be selected as the target historical content. For example, if the quantity difference is 3, the historical hot content ranked in the top three will be determined as the target historical content.

[0085] The beneficial effect of such a setting is that historical hot content with high views is selected as the target historical content, which increases the user's interest in watching live broadcasts, avoids reporting content that the user does not care about, improves the accuracy of content reporting, and enhances the user experience.

[0086] In this embodiment, the method further includes: obtaining a preset date database; wherein the preset date database includes knowledge content corresponding to each date; and obtaining knowledge content corresponding to the current date from the preset date database as the content to be broadcast.

[0087] Specifically, a date database may be pre-set, and the date database may store knowledge content corresponding to each date, for example, relevant content of each solar term, festival, anniversary, etc.

[0088] Determine the current date, and search the date database for knowledge content corresponding to the current date, that is, determine whether the current day belongs to any solar term, festival, anniversary, etc., so as to determine the knowledge related to the current day as the content to be broadcast. One or more relevant knowledge contents can be searched from the date database as the content to be broadcast.

[0089] The beneficial effect of this setting is that relevant knowledge of the day is selected to improve the accuracy of determining the content to be broadcast. Through the content database, history database, and date database, multi-level mining of the content to be broadcast is achieved, which enriches the live broadcast content and enhances the viewing interest of users.

[0090] S303: extracting live broadcast information from the content to be broadcast; wherein the live broadcast information represents key information in the content to be broadcast.

[0091] Exemplarily, this step may refer to the above-mentioned step S102 and will not be described in detail.

[0092] S304: Based on the preset digital human image, broadcast live information in a pre-built live broadcast room.

[0093] Exemplarily, this step may refer to the above-mentioned step S103 and will not be described in detail.

[0094] In the disclosed embodiment, the content information of the event to be broadcast, i.e., the content to be broadcast, is automatically determined from the preset content database, and there is no need to manually confirm what event needs to be broadcast, which reduces manual operations and improves the efficiency of live broadcasting. The content to be broadcasted is automatically processed into live broadcast information that can be displayed in the live broadcast room, enriching the form of data played in the live broadcast room, further saving manpower and time, and improving the fun of the broadcast. Live broadcasting through the preset digital human image makes the live broadcast more vivid and enhances the user's viewing experience.

[0095] Figure 4A flowchart of a digital human-based live broadcast method provided in an embodiment of the present disclosure.

[0096] In this embodiment, extracting live broadcast information from the content to be broadcast includes: segmenting the content to be broadcast to obtain word segments in the content to be broadcast; determining extended content of the content to be broadcast according to the word segments in the content to be broadcast; wherein the extended content represents content related to the word segments in the content to be broadcast; and extracting live broadcast information from the content to be broadcast and the extended content.

[0097] like Figure 4 As shown, the method comprises the following steps:

[0098] S401, determining content to be broadcast from a preset content database; wherein the content to be broadcast represents content information of an event to be broadcast.

[0099] Exemplarily, this step may refer to the above-mentioned step S101 and will not be described in detail.

[0100] S402: segment the content to be broadcast to obtain segmented words in the content to be broadcast.

[0101] For example, after obtaining the content to be broadcast, the live broadcast information can be determined according to the content to be broadcast. The content to be broadcast can be segmented first, and the content to be broadcast can be segmented into multiple segmented words. The segmented segmented words can be entity words with specific meanings, or sentence words, auxiliary words, etc. without specific meanings. In this embodiment, the segmentation method is not specifically limited.

[0102] S403: Determine extended content of the content to be broadcast according to the segmented words in the content to be broadcast; wherein the extended content represents content related to the segmented words in the content to be broadcast.

[0103] Exemplarily, associations between different words and contents are preset and stored. For example, if the word is "tiger", then tiger-related contents are associated. According to the preset association, the contents associated with the segmented words in the content to be broadcast are determined as extended contents. One content to be broadcast may correspond to multiple segmented words, and one segmented word may correspond to multiple extended contents, that is, one content to be broadcast may correspond to multiple extended contents.

[0104] The segmented words of the content to be broadcast can be filtered, and it is only necessary to determine the extended content corresponding to the segmented words retained by the screening. For example, modal particles and auxiliary words can be filtered out, and entity words can be retained.

[0105] In this embodiment, the extended content of the content to be broadcast is determined based on the segmented words in the content to be broadcast, including: performing natural language processing on the segmented words in the content to be broadcast to obtain the semantic score of the segmented words in the content to be broadcast; wherein the semantic score represents the semantic importance of the segmented words; according to the semantic score of the segmented words in the content to be broadcast, determining the target segmented words from the segmented words in the content to be broadcast; and determining the extended content of the content to be broadcast based on the target segmented words.

[0106] Specifically, for the segmented words in the content to be broadcast, a preset NLP (Natural Language Processing) tool can be used to perform natural language processing on the segmented words in the content to be broadcast. For example, semantic recognition can be performed on the segmented words, and the semantic score of the segmented words in the content to be broadcast can be obtained according to the semantics of the segmented words. The semantic score can represent the semantic importance of the segmented words. The more important the segmented words are, the higher the semantic score is. For example, the semantic scores of modal particles and auxiliary words are relatively low, and the semantic scores of entity words are relatively high.

[0107] According to the semantic scores of the segmented words in the content to be broadcasted, the segmented words in the content to be broadcasted are sorted, and according to the sorting results, the target segmented words are determined from the segmented words in the content to be broadcasted. For example, the segmented words can be sorted in descending order according to the semantic scores, and the first K segmented words are determined as the target segmented words. According to the target segmented words, the extended content of the content to be broadcasted is determined. For example, the target segmented words can be expanded, and each target segmented word can be expanded into a complete sentence as the extended content. Alternatively, each target segmented word can be combined into a complete sentence as the extended content.

[0108] The beneficial effect of such a setting is that the segmented words are scored and screened, important segmented words are expanded, the efficiency and accuracy of the expansion are improved, and the efficiency and accuracy of determining the live broadcast information are improved.

[0109] In this embodiment, determining the extended content of the content to be broadcast according to the target segmentation includes: determining the content corresponding to the target segmentation as the extended content of the content to be broadcast according to the association relationship between the preset segmentation and the content.

[0110] Specifically, a correlation relationship between a segmentation and content is preset, and after determining a target segmentation, the content corresponding to the target segmentation is searched according to the correlation relationship as the extended content of the content to be broadcasted. If there are multiple target segmentations, multiple extended contents can be obtained.

[0111] The association between word segmentation and content can be represented in the form of a knowledge graph. For example, a word segmentation can be used as a node in the knowledge graph, and a piece of content can also be used as a node in the knowledge graph. The node of the word segmentation can be connected to the node of the content, indicating that there is an association between the word segmentation and the content. Nodes of different word segmentations can also be connected by edges, and nodes of different content can also be connected by edges. When determining the extended content, you can find content that has a direct edge connection with the node of the target word segmentation as the extended content.

[0112] The beneficial effect of such a setting is that, by using a preset association relationship, the extended content is determined, which effectively improves the efficiency of determining the extended content, thereby improving the efficiency of determining the live broadcast information.

[0113] S404: extracting live broadcast information from the content to be broadcast and the extended content.

[0114] Exemplarily, the live broadcast information can be determined based on the content to be broadcast and the extended content in combination. For example, the content to be broadcast and the extended content can be spliced ​​first, and then semantic recognition can be performed on the spliced ​​content, and key information can be extracted from the spliced ​​content, and the extracted key information can be determined as the live broadcast information. The key information can represent the meaning of the spliced ​​content, for example, a paragraph or a sentence can be extracted from the spliced ​​content, or the spliced ​​content can be summarized.

[0115] You can also combine the content to be broadcast, extended content, and the current date to get live broadcast information, thereby improving the comprehensiveness of the live broadcast information.

[0116] In this embodiment, the content to be broadcast is expanded, the accuracy of determining the live broadcast information is improved, the content to be played in the live broadcast room is enriched, the interest of the broadcast is increased, and the user experience is improved.

[0117] In this embodiment, live broadcast information is extracted from the content to be broadcast and the extended content, including: generating prompt words based on a preset prompt word template according to the content to be broadcast and the extended content; wherein the prompt words represent the content to be broadcast and the extended content; inputting the prompt words into a preset large model to obtain live broadcast information; wherein the preset large model is a neural network model, which is used to output live broadcast information according to the prompt words, and the live broadcast information includes at least one of a title, a text, and a background picture of a live broadcast room.

[0118] Specifically, a large model is pre-built, which is a neural network model that can output live broadcast information. When the large model is working, it can process data according to prompt words. The prompt words are pre-set text data that can assist the large model to better understand the task. The prompt words are input into the large model to obtain live broadcast information.

[0119] A prompt word template is pre-set. According to the content to be broadcast and the extended content, the prompt word template is supplemented to obtain a complete prompt word, that is, the prompt word can represent the content to be broadcast and the extended content. The prompt word is input into the preset large model to obtain the live broadcast information. The live broadcast information may include a title, a copy, a background picture of the live broadcast room, etc. Different prompt word templates can be preset for different live broadcast information, that is, different prompt words can be generated. For example, a prompt word template specifically used to generate a copy can be preset, and the prompt word template can be "reference information includes mined hot data, today's holiday information, and extended reference information; the mined hot data is XXX, today's holiday information is XXX, and the extended reference information is XXX; this is a live copy generation task, generating content suitable for the anchor to broadcast, and the specific requirements are as follows: obtain content from the reference information, the context is coherent and smooth, the content is healthy and positive, and there is no conflict with the reference information. Commentary content can be appropriately added, and the number of words is controlled at about 500 words." Among them, the mined hot data can be the content to be broadcast, today's holiday information can be the current date, and the extended reference information can be the extended content.

[0120] The live broadcast information can also include information in the form of questions and answers. One question can correspond to three options, of which only one option is the correct answer. Through the preset large model, relevant knowledge can be extracted from the prompt words to generate question and answer questions. The prompt word template for generating question and answer questions can be "reference information includes mined hot data, today's holiday information, and extended reference information; the mined hot data is XXX, today's holiday information is XXX, and the extended reference information is XXX; this is a question and answer generation task, one question, three options, of which only one option is correct, and the other options are close to the correct option. Note: questions and options should be as concise as possible, with questions controlled to about 20 words and options controlled to within 5 words."

[0121] The beneficial effect of this setting is that titles, scripts, backgrounds, etc. are automatically generated through large models, saving manpower and time, increasing the richness of the content displayed in the live broadcast, making it easier to attract users and improving user experience.

[0122] S405: Based on the preset digital human image, broadcast live information in a pre-built live broadcast room.

[0123] Exemplarily, the digital human live broadcast room synthesis technology is used to synthesize the live broadcast room, input the copy script to be broadcast, the background material in the live broadcast room, the texture material, the question and answer questions, etc., and drive the digital human in the live broadcast room to broadcast each live broadcast information in turn.

[0124] Figure 5 This is a schematic diagram of the live broadcast room. Figure 5There are questions and answers displayed in the video, and the digital human anchor can read the questions and answers during the live broadcast. Users can choose answers by clicking on the options on the screen to achieve live broadcast interaction and increase the fun of the live broadcast.

[0125] In the disclosed embodiment, the content information of the event to be broadcast, i.e., the content to be broadcast, is automatically determined from the preset content database, and there is no need to manually confirm what event needs to be broadcast, which reduces manual operations and improves the efficiency of live broadcasting. The content to be broadcasted is automatically processed into live broadcast information that can be displayed in the live broadcast room, enriching the form of data played in the live broadcast room, further saving manpower and time, and improving the fun of the broadcast. Live broadcasting through the preset digital human image makes the live broadcast more vivid and enhances the user's viewing experience.

[0126] Figure 6 This is a structural block diagram of a digital human-based live broadcast device provided in an embodiment of the present disclosure. For ease of explanation, only the parts related to the embodiment of the present disclosure are shown. Figure 6 The digital human-based live broadcast device 600 includes: a content determination unit 601, an information extraction unit 602 and a live broadcast unit 603.

[0127] The content determination unit 601 is used to determine the content to be broadcast from a preset content database; wherein the content to be broadcast represents the content information of the event to be broadcast;

[0128] The information extraction unit 602 is used to extract live broadcast information from the content to be broadcast; wherein the live broadcast information represents key information in the content to be broadcast;

[0129] The live broadcast unit 603 is used to play the live broadcast information in a pre-built live broadcast room based on the preset digital human image.

[0130] Figure 7 A structural block diagram of a digital human-based live broadcast device provided in an embodiment of the present disclosure, such as Figure 7 As shown, the digital human-based live broadcast device 700 includes a content determination unit 701, an information extraction unit 702 and a live broadcast unit 703, wherein the preset content database includes current hot content, and the current hot content represents the content information of the current hot event; the content determination unit 701 includes a pageview acquisition module 7011 and a content determination module 7012.

[0131] The pageview acquisition module 7011 is used to acquire pageview information of the current hot content in the preset content database;

[0132] The content determination module 7012 is used to determine the content to be broadcast from the preset content database according to the pageview information of each current hot content in the preset content database.

[0133] In one example, the content determination module 7012 includes:

[0134] A level determination submodule, used to determine the content level of the current hot content according to the pageview information of the current hot content in the preset content database; wherein the content level represents the influence of the content;

[0135] The content determination submodule is used to determine the current hot content at a preset content level as the content to be broadcast.

[0136] In one example, the content determination submodule is specifically used to:

[0137] If the number of current hot content at the preset content level is less than a preset number threshold, a preset historical database is obtained; wherein the preset historical database includes historical hot content, and the historical hot content represents content information of hot events occurring within a preset historical time period;

[0138] Determining target historical content from historical hot content in the preset historical database according to the difference between the quantity of the current hot content at the preset content level and the preset quantity threshold;

[0139] The current hot content and the target historical content at a preset content level are determined as the content to be broadcast.

[0140] In one example, the content determination submodule is specifically used to:

[0141] Obtaining the pageview information of the historical hot content in the preset historical database;

[0142] Sorting the historical hot spots according to the pageview information of each historical hot spot content in the preset historical database;

[0143] The target historical content is determined according to the sorting result and the quantity difference between the quantity of the current hot content at the preset content level and the preset quantity threshold.

[0144] In one example, it also includes:

[0145] A database acquisition unit, used to acquire a preset date database; wherein the preset date database includes knowledge content corresponding to each date;

[0146] The knowledge acquisition unit is used to acquire the knowledge content corresponding to the current date from the preset date database as the content to be broadcast.

[0147] In one example, the information extraction unit 702 includes:

[0148] A segmentation module, used for segmenting the content to be broadcast to obtain segmented words in the content to be broadcast;

[0149] An extension module, used to determine the extended content of the content to be broadcast according to the segmented words in the content to be broadcast; wherein the extended content represents the content related to the segmented words in the content to be broadcast;

[0150] An extraction module is used to extract the live broadcast information from the content to be broadcast and the extended content.

[0151] In one example, an extension module includes:

[0152] A semantic determination submodule, used for performing natural language processing on the segmented words in the content to be broadcast, and obtaining the semantic score of the segmented words in the content to be broadcast; wherein the semantic score represents the semantic importance of the segmented words;

[0153] A target determination submodule, configured to determine a target segmented word from the segmented words in the content to be broadcast according to the semantic scores of the segmented words in the content to be broadcast;

[0154] The content extension submodule is used to determine the extended content of the content to be broadcast according to the target word segmentation.

[0155] In one example, the content extension submodule is specifically used to:

[0156] According to the preset association relationship between the segmented words and the content, the content corresponding to the target segmented word is determined as the extended content of the content to be broadcasted.

[0157] In one example, the extraction module includes:

[0158] A prompt word generation submodule, used to generate prompt words according to the content to be broadcast and the extended content based on a preset prompt word template; wherein the prompt words represent the content to be broadcast and the extended content;

[0159] The information acquisition submodule is used to input the prompt word into a preset large model to obtain the live broadcast information; wherein the preset large model is a neural network model, which is used to output the live broadcast information according to the prompt word, and the live broadcast information includes at least one of the title, text, and background picture of the live broadcast room.

[0160] In one example, the live broadcast unit 703 includes:

[0161] The sequential playback module is used to play the live broadcast information corresponding to the to-be-played content in sequence in the pre-built live broadcast room according to the storage order of the to-be-played content in the preset content database if multiple to-be-played content is determined from the preset content database.

[0162] In one example, it also includes:

[0163] A question acquisition unit, used to acquire question information sent by the user in the live broadcast room;

[0164] A question processing unit, used to perform semantic recognition processing on the question information to obtain answer information corresponding to the question information;

[0165] The reply unit is used to play the reply information in the pre-built live broadcast room based on the preset image of the digital human.

[0166] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device.

[0167] Figure 8 A structural block diagram of an electronic device provided in an embodiment of the present disclosure, such as Figure 8 As shown, the electronic device 800 includes: at least one processor 802; and a memory 801 communicatively connected to the at least one processor 802; wherein the memory stores instructions executable by the at least one processor 802, and the instructions are executed by the at least one processor 802 so that the at least one processor 802 can execute the digital human-based live broadcast method disclosed in the present invention.

[0168] The electronic device 800 further includes a receiver 803 and a transmitter 804. The receiver 803 is used to receive instructions and data sent by other devices, and the transmitter 804 is used to send instructions and data to external devices.

[0169] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0170] According to an embodiment of the present disclosure, the present disclosure also provides a computer program product, which includes: a computer program, the computer program is stored in a readable storage medium, at least one processor of an electronic device can read the computer program from the readable storage medium, and at least one processor executes the computer program so that the electronic device executes the solution provided by any of the above embodiments.

[0171] Fig. 9A schematic block diagram of an example electronic device 900 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0172] like Fig. 9 As shown, the device 900 includes a computing unit 901, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 902 or a computer program loaded from a storage unit 908 into a random access memory (RAM) 903. In the RAM 903, various programs and data required for the operation of the device 900 can also be stored. The computing unit 901, the ROM 902, and the RAM 903 are connected to each other via a bus 904. An input / output (I / O) interface 905 is also connected to the bus 904.

[0173] A number of components in the device 900 are connected to the I / O interface 905, including: an input unit 906, such as a keyboard, a mouse, etc.; an output unit 907, such as various types of displays, speakers, etc.; a storage unit 908, such as a disk, an optical disk, etc.; and a communication unit 909, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 909 allows the device 900 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0174] The computing unit 901 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 901 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 901 performs the various methods and processes described above, such as a live broadcast method based on a digital person. For example, in some embodiments, the live broadcast method based on a digital person may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 908. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 900 via the ROM 902 and / or the communication unit 909. When the computer program is loaded into the RAM 903 and executed by the computing unit 901, one or more steps of the live broadcast method based on a digital person described above may be performed. Alternatively, in other embodiments, the computing unit 901 may be configured to perform a live broadcast method based on a digital person in any other appropriate manner (e.g., by means of firmware).

[0175] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0176] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0177] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0178] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0179] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0180] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services ("Virtual Private Server", or "VPS" for short). The server may also be a server of a distributed system, or a server combined with a blockchain.

[0181] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0182] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A live broadcast method based on digital human, comprising: Determine the content to be broadcast from a preset content database; wherein the content to be broadcast represents content information of the event to be broadcast; Extracting live broadcast information from the content to be broadcast; wherein the live broadcast information represents key information in the content to be broadcast; Based on the preset digital human image, the live broadcast information is played in a pre-built live broadcast room.

2. The method according to claim 1, wherein: The preset content database includes current hot content, and the current hot content represents content information of the current hot event; Determining the content to be broadcast from a preset content database includes: Obtaining the pageview information of the current hot content in the preset content database; The content to be broadcast is determined from the preset content database according to the pageview information of each current hot content in the preset content database.

3. The method according to claim 2, wherein: The determining the content to be broadcast from the preset content database according to the pageview information of each current hot content in the preset content database comprises: Determine the content level of the current hot content according to the pageview information of the current hot content in the preset content database; wherein the content level represents the influence of the content; The current hot content at a preset content level is determined as the content to be broadcast.

4. The method according to claim 3, wherein: The step of determining the current hot content at a preset content level as the content to be broadcast includes: If the number of current hot content at the preset content level is less than a preset number threshold, a preset historical database is obtained; wherein the preset historical database includes historical hot content, and the historical hot content represents content information of hot events occurring within a preset historical time period; Determining target historical content from historical hot content in the preset historical database according to the difference between the quantity of the current hot content at the preset content level and the preset quantity threshold; The current hot content and the target historical content at a preset content level are determined as the content to be broadcast.

5. The method according to claim 4, wherein: The step of determining the target historical content from the historical hot content in the preset historical database according to the quantity difference between the quantity of the current hot content at the preset content level and the preset quantity threshold comprises: Obtaining the pageview information of the historical hot content in the preset historical database; Sorting the historical hot spots according to the pageview information of each historical hot spot content in the preset historical database; The target historical content is determined according to the sorting result and the quantity difference between the quantity of the current hot content at the preset content level and the preset quantity threshold.

6. The method according to any one of claims 1 to 5, further comprising: Acquire a preset date database; wherein the preset date database includes knowledge content corresponding to each date; The knowledge content corresponding to the current date is obtained from the preset date database as the content to be broadcast.

7. The method according to any one of claims 1 to 6, wherein: The extracting live broadcast information from the content to be broadcast includes: Segmenting the content to be broadcast to obtain segmented words in the content to be broadcast; Determine the extended content of the content to be broadcast according to the segmented words in the content to be broadcast; wherein the extended content represents the content related to the segmented words in the content to be broadcast; The live broadcast information is extracted from the content to be broadcast and the extended content.

8. The method according to claim 7, wherein: The step of determining the extended content of the content to be broadcast according to the segmented words in the content to be broadcast comprises: Performing natural language processing on the segmented words in the content to be broadcast to obtain semantic scores of the segmented words in the content to be broadcast; wherein the semantic scores represent the semantic importance of the segmented words; Determining a target segmented word from the segmented words in the content to be broadcast according to the semantic scores of the segmented words in the content to be broadcast; According to the target segmentation, extended content of the content to be broadcast is determined.

9. The method according to claim 8, wherein: The step of determining the extended content of the content to be broadcast according to the target word segmentation includes: According to the preset association relationship between the segmented words and the content, the content corresponding to the target segmented word is determined as the extended content of the content to be broadcasted.

10. The method according to any one of claims 7 to 9, wherein: The extracting the live broadcast information from the content to be broadcast and the extended content includes: Generate prompt words according to the content to be broadcast and the extended content based on a preset prompt word template; wherein the prompt words represent the content to be broadcast and the extended content; The prompt word is input into a preset large model to obtain the live broadcast information; wherein the preset large model is a neural network model, which is used to output the live broadcast information according to the prompt word, and the live broadcast information includes at least one of the title, text, and background picture of the live broadcast room.

11. The method according to any one of claims 1 to 10, wherein: Playing the live broadcast information in a pre-built live broadcast room includes: If a plurality of contents to be broadcast are determined from the preset content database, the live broadcast information corresponding to the contents to be broadcast are played in sequence in the pre-built live broadcast room according to the storage order of the contents to be broadcast in the preset content database.

12. The method according to any one of claims 1 to 11, further comprising: Get the question information sent by the user in the live broadcast room; Performing semantic recognition processing on the question information to obtain answer information corresponding to the question information; Based on the preset image of the digital human, the reply information is played in the pre-built live broadcast room.

13. A live broadcast device based on digital human, comprising: A content determination unit, configured to determine the content to be broadcast from a preset content database; wherein the content to be broadcast represents content information of the event to be broadcast; An information extraction unit, configured to extract live broadcast information from the content to be broadcast; wherein the live broadcast information represents key information in the content to be broadcast; The live broadcast unit is used to play the live broadcast information in a pre-built live broadcast room based on the preset digital human image.

14. The device according to claim 13, wherein: The preset content database includes current hot content, and the current hot content represents content information of the current hot event; The content determination unit includes: A pageview acquisition module, used to acquire pageview information of current hot content in the preset content database; The content determination module is used to determine the content to be broadcast from the preset content database according to the pageview information of each current hot content in the preset content database.

15. The device according to claim 14, wherein: The content determination module comprises: A level determination submodule, used to determine the content level of the current hot content according to the pageview information of the current hot content in the preset content database; wherein the content level represents the influence of the content; The content determination submodule is used to determine the current hot content at a preset content level as the content to be broadcast.

16. The device according to claim 15, wherein: The content determination submodule is specifically used for: If the number of current hot content at the preset content level is less than a preset number threshold, a preset historical database is obtained; wherein the preset historical database includes historical hot content, and the historical hot content represents content information of hot events occurring within a preset historical time period; Determining target historical content from historical hot content in the preset historical database according to the difference between the quantity of the current hot content at the preset content level and the preset quantity threshold; The current hot content and the target historical content at a preset content level are determined as the content to be broadcast.

17. The device according to claim 16, wherein: The content determination submodule is specifically used for: Obtaining the pageview information of the historical hot content in the preset historical database; Sorting the historical hot spots according to the pageview information of each historical hot spot content in the preset historical database; The target historical content is determined according to the sorting result and the quantity difference between the quantity of the current hot content at the preset content level and the preset quantity threshold.

18. The apparatus according to any one of claims 13 to 17, further comprising: A database acquisition unit, used to acquire a preset date database; wherein the preset date database includes knowledge content corresponding to each date; The knowledge acquisition unit is used to acquire the knowledge content corresponding to the current date from the preset date database as the content to be broadcast.

19. The device according to any one of claims 13 to 18, wherein: The information extraction unit comprises: A segmentation module, used for segmenting the content to be broadcast to obtain segmented words in the content to be broadcast; An extension module, used to determine the extended content of the content to be broadcast according to the segmented words in the content to be broadcast; wherein the extended content represents the content related to the segmented words in the content to be broadcast; An extraction module is used to extract the live broadcast information from the content to be broadcast and the extended content.

20. The device according to claim 19, wherein The expansion module comprises: A semantic determination submodule, used for performing natural language processing on the segmented words in the content to be broadcast, and obtaining the semantic score of the segmented words in the content to be broadcast; wherein the semantic score represents the semantic importance of the segmented words; A target determination submodule, configured to determine a target segmented word from the segmented words in the content to be broadcast according to the semantic scores of the segmented words in the content to be broadcast; The content extension submodule is used to determine the extended content of the content to be broadcast according to the target word segmentation.

21. The device according to claim 20, wherein: The content extension submodule is specifically used for: According to the preset association relationship between the segmented words and the content, the content corresponding to the target segmented word is determined as the extended content of the content to be broadcasted.

22. The device according to any one of claims 19 to 21, wherein: The extraction module comprises: A prompt word generation submodule, used to generate prompt words according to the content to be broadcast and the extended content based on a preset prompt word template; wherein the prompt words represent the content to be broadcast and the extended content; The information acquisition submodule is used to input the prompt word into a preset large model to obtain the live broadcast information; wherein the preset large model is a neural network model, which is used to output the live broadcast information according to the prompt word, and the live broadcast information includes at least one of the title, text, and background picture of the live broadcast room.

23. The device according to any one of claims 13 to 22, wherein: The live broadcast unit comprises: The sequential playback module is used to play the live broadcast information corresponding to the to-be-played content in sequence in the pre-built live broadcast room according to the storage order of the to-be-played content in the preset content database if multiple to-be-played content is determined from the preset content database.

24. The apparatus according to any one of claims 13 to 23, further comprising: A question acquisition unit, used to acquire question information sent by the user in the live broadcast room; A question processing unit, used to perform semantic recognition processing on the question information to obtain answer information corresponding to the question information; The reply unit is used to play the reply information in the pre-built live broadcast room based on the preset image of the digital human.

25. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 12.

26. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-12.

27. A computer program product, wherein: The invention comprises a computer program, which implements the steps of the method according to any one of claims 1 to 12 when being executed by a processor.

Images