CN112287173A

CN112287173A - Method and apparatus for generating information

Info

Publication number: CN112287173A
Application number: CN202011196608.8A
Authority: CN
Inventors: 张同新; 陈婉君; 姚佳立; 张昊宇
Original assignee: Beijing Youzhuju Network Technology Co Ltd
Current assignee: Beijing Youzhuju Network Technology Co Ltd
Priority date: 2020-10-30
Filing date: 2020-10-30
Publication date: 2021-01-29

Abstract

The embodiment of the disclosure discloses a method and a device for generating information. One embodiment of the method comprises: acquiring a keyword; acquiring a candidate information set corresponding to the keyword, and acquiring a content tag of candidate information in the candidate information set; for candidate information in the candidate information set, determining an index value of the candidate information according to the content tag and the keyword of the candidate information; selecting candidate information from the candidate information set as an information material according to index values corresponding to the candidate information in the candidate information set respectively; and generating target information according to the information material. This embodiment contributes to reducing the generation cycle of the target information.

Description

Method and apparatus for generating information

Technical Field

Embodiments of the present disclosure relate to the field of computer technologies, and in particular, to a method and an apparatus for generating information.

Background

With the development of the application of the internet, each user in the internet can show or provide own works to other users through an internet platform. The forms of works authored by users are also diverse and include video, audio, images, text, and the like. The internet also enables the accumulation, re-creation, and dissemination of information through the display and sharing of the user's works.

Most of the information displayed and shared in the internet is produced through a process that a user selects proper information from massive information provided by the internet as a material according to the design idea of the user, and then the information is adjusted based on the selected material to generate the information meeting the design idea of the user.

For example, when a user makes a video, the user usually needs to find a video clip that matches the subject of the video he wants to make, and may also need to find a suitable audio as a dubbing of the video, and then integrate the video clip and the audio to generate a new video.

In the information production process, because the information provided by the current internet is massive, a lot of time is usually spent on screening materials, and the quality of the selected materials can directly influence the quality and the presentation effect of the finally generated video.

Disclosure of Invention

Embodiments of the present disclosure propose methods and apparatuses for generating information.

In a first aspect, an embodiment of the present disclosure provides a method for generating information, the method including: acquiring a keyword; acquiring a candidate information set corresponding to the keyword, and acquiring a content tag of candidate information in the candidate information set; for candidate information in the candidate information set, determining an index value of the candidate information according to the content tag and the keyword of the candidate information; selecting candidate information from the candidate information set as an information material according to index values corresponding to the candidate information in the candidate information set respectively; and generating target information according to the information material.

In a second aspect, an embodiment of the present disclosure provides an apparatus for generating information, the apparatus including: a first acquisition unit configured to acquire a keyword; a second acquisition unit configured to acquire a candidate information set corresponding to the keyword, and acquire a content tag of candidate information in the candidate information set; a determination unit configured to determine, for candidate information in the candidate information set, an index value of the candidate information according to a content tag and a keyword of the candidate information; the selection unit is configured to select the candidate information from the candidate information set as an information material according to the index values respectively corresponding to the candidate information in the candidate information set; a generating unit configured to generate target information from the information material.

In a third aspect, an embodiment of the present disclosure provides an electronic device, including: one or more processors; storage means for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the method as described in any implementation of the first aspect.

In a fourth aspect, embodiments of the present disclosure provide a computer-readable medium on which a computer program is stored, which computer program, when executed by a processor, implements the method as described in any of the implementations of the first aspect.

According to the method and the device for generating the information, the candidate information set corresponding to the keyword and the content label of each candidate information in the candidate information set are obtained, the index value of each candidate information is determined according to the content label of each candidate information and the obtained keyword, and the candidate information is selected from the candidate information set as the information material according to the corresponding index value to generate the target information. Therefore, the information material of the target information expected to be generated can be quickly acquired from the mass information, and the generation period of the target information is reduced.

Drawings

Other features, objects and advantages of the disclosure will become more apparent upon reading of the following detailed description of non-limiting embodiments thereof, made with reference to the accompanying drawings in which:

FIG. 1 is an exemplary system architecture diagram in which one embodiment of the present disclosure may be applied;

FIG. 2 is a flow diagram for one embodiment of a method for generating information, according to the present disclosure;

FIG. 3 is a flow diagram of yet another embodiment of a method for generating information according to the present disclosure;

FIG. 4 is a schematic diagram of one application scenario of a method for generating information in accordance with an embodiment of the present disclosure;

FIG. 5 is a schematic block diagram illustrating one embodiment of an apparatus for generating information according to the present disclosure;

FIG. 6 is a schematic structural diagram of an electronic device suitable for use in implementing embodiments of the present disclosure.

Detailed Description

The present disclosure is described in further detail below with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the relevant invention and not restrictive of the invention. It should be noted that, for convenience of description, only the portions related to the related invention are shown in the drawings.

It should be noted that, in the present disclosure, the embodiments and features of the embodiments may be combined with each other without conflict. The present disclosure will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.

Fig. 1 illustrates an exemplary architecture 100 to which embodiments of the disclosed method for generating information or apparatus for generating information may be applied.

As shown in fig. 1, the system architecture 100 may include

terminal devices

101, 102, 103, a network 104, and a server 105. The network 104 serves as a medium for providing communication links between the

terminal devices

101, 102, 103 and the server 105. Network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, to name a few.

The

terminal devices

101, 102, 103 interact with a server 105 via a network 104 to receive or send messages or the like. Various client applications may be installed on the

terminal devices

101, 102, 103. Such as browser-like applications, search-like applications, shopping-like applications, instant messaging tools, social platform applications, information flow-like applications, and so forth.

The

terminal apparatuses

101, 102, and 103 may be hardware or software. When the

terminal devices

101, 102, 103 are hardware, they may be various electronic devices including, but not limited to, smart phones, tablet computers, e-book readers, laptop portable computers, desktop computers, and the like. When the

terminal apparatuses

101, 102, 103 are software, they can be installed in the electronic apparatuses listed above. It may be implemented as multiple pieces of software or software modules (e.g., multiple pieces of software or software modules to provide distributed services) or as a single piece of software or software module. And is not particularly limited herein.

The server 105 may be a server providing various services, such as a server providing back-end support for client applications installed on the

terminal devices

101, 102, 103. The server 105 may generate target information based on the acquired keywords and transmit the target information to the

terminal apparatuses

101, 102, and 103 to be presented on the

terminal apparatuses

101, 102, and 103.

It should be noted that the process of the server 105 generating the target information based on the obtained keywords may not need to interact with the

terminal devices

101, 102, and 103. At this time, the

terminal apparatuses

101, 102, 103 and the network 104 may not exist.

It should be noted that the method for generating information provided by the embodiment of the present disclosure is generally performed by the server 105, and accordingly, the apparatus for generating information is generally disposed in the server 105.

It should be noted that the

terminal devices

101, 102, and 103 may also generate the target information with the acquired keywords, in this case, the method for generating the information may be executed by the

terminal devices

101, 102, and 103, and accordingly, the apparatus for generating the information may also be provided in the

terminal devices

101, 102, and 103. At this point, the exemplary system architecture 100 may not have the server 105 and the network 104.

The server 105 may be hardware or software. When the server 105 is hardware, it may be implemented as a distributed server cluster composed of a plurality of servers, or may be implemented as a single server. When the server 105 is software, it may be implemented as multiple pieces of software or software modules (e.g., multiple pieces of software or software modules used to provide distributed services), or as a single piece of software or software module. And is not particularly limited herein.

It should be understood that the number of terminal devices, networks, and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.

With continued reference to FIG. 2, a flow 200 of one embodiment of a method for generating information in accordance with the present disclosure is shown. The method for generating information comprises the following steps:

step 201, keywords are obtained.

In this embodiment, the keywords may be predetermined according to actual application requirements and application scenarios. In general, the keywords may be predetermined according to target information desired to be generated. In this case, the keyword may be a keyword of the target information desired to be generated. The target information may be various types of information, among others. Types of target information include, but are not limited to, at least one of: text, image, video, audio.

As an example, when the target information is an introduction video of a certain person, the name of the person and words for describing contents specifically desired to be introduced in the video may be determined as keywords.

In the present embodiment, the execution subject of the method for generating information (e.g., server 105 shown in fig. 1) may acquire the keyword from a local or other storage device (e.g., a terminal device used by the designer of the target information, etc.).

As an example, a keyword may be determined by a designer of target information according to contents included in the target information desired by the designer, and then transmitted to the execution subject.

Step 202, a candidate information set corresponding to the keyword is obtained, and a content tag of the candidate information in the candidate information set is obtained.

In the present embodiment, the candidate information may also be various types of information. The type of candidate information includes, but is not limited to, at least one of: text, image, video, audio. The type of the candidate information may be the same as or different from that of the target information. The candidate information corresponding to the keyword may refer to candidate information related to the keyword.

The content tag of the candidate information may be used to describe the content of the candidate information. For example, when the candidate information is a video, the content tag may be used to describe an object appearing in the video (such as an attribute of the object, for example), and may also be used to describe an event in the video (such as a keyword of the event, for example). It should be understood that for a candidate message, there may be one or more content tags of the candidate message.

In this embodiment, a correspondence between the candidate information and the keyword and a correspondence between the candidate information and the content tag may be established in advance. At this time, the executing body may acquire candidate information corresponding to the keyword to form a candidate information set by using the pre-established correspondence, and acquire a content tag corresponding to each candidate information in the candidate information set.

Step 203, for the candidate information in the candidate information set, determining the index value of the candidate information according to the content label and the keyword of the candidate information.

In the embodiment, the index value may be an index value of various indexes according to an actual application requirement and different application scenarios. In general, an index value may be used to reflect the attributes of a candidate information in some respect. For example, the index value may be used to characterize the click volume of the candidate information. For another example, the index value may be used to characterize a degree of matching between the candidate information and the target information desired to be generated.

It should be appreciated that the indicator value of the candidate information may reflect one or more attributes of the candidate information. For example, the index value may characterize the click amount of the candidate information, and may also characterize the matching degree between the candidate information and the target information desired to be generated.

In this embodiment, according to the different meanings indicated by the index values and the actual application requirements, various methods can be flexibly adopted to determine the index values of the candidate information according to the content tags of the candidate information and the acquired keywords.

For example, for a candidate information, the similarity between the content tag of the candidate information and the acquired keyword may be calculated as the index value of the candidate information. For another example, for a candidate information, a similarity between the candidate information and the obtained keyword may be calculated first, then a similarity between the content tag of the candidate information and the obtained keyword may be calculated, and then an average of the two similarities may be calculated as the index value of the candidate information.

And step 204, selecting the candidate information from the candidate information set as an information material according to the index values corresponding to the candidate information in the candidate information set.

In this embodiment, according to different application requirements, candidate information can be flexibly selected from the candidate information set as an information material. Wherein, the information material may refer to a material of the target information desired to be generated.

For example, a preset number of candidate information may be selected from the candidate information set as the information material in the order of the index value from large to small or from small to large. For another example, the candidate information with the index value larger than the preset index value threshold may be selected from the candidate information set as the information material. For another example, the candidate information set may be selected from the candidate information set, wherein the index value is greater than the threshold value, and then the selected candidate information set may be selected as the information material.

Step 205, generating target information according to the information material.

In this embodiment, the target information may be generated using information material. For example, the information material may be adjusted to actual needs to generate the target information. It should be appreciated that the target information may be generated in a variety of different ways depending on the type of target information, the type of information material, and so forth.

For example, the type of information material and the type of target information are both audio. At this time, a desired audio can be generated using the audio included in the information material by using various existing audio production methods. For another example, the type of the information material is an image, and the type of the target information is a video. In this case, a desired video can be generated using images included in the information material by using various existing video production methods.

In some optional implementation manners of this embodiment, the description information of the target information may be obtained first, and then the keyword may be determined according to the obtained description information. The description information of the object information may include various information for describing the object information. For example, the description information may include, but is not limited to, information for describing the content, category, duration, and the like of the target information.

As an example, when the target information is an image, the description information may be used to describe the content included in the image desired to be generated. When the target information is a video, the description information may be used to describe a scene that the video desired to be generated includes.

Generally, the description information may be determined in advance by a technician such as a designer of the target information and then transmitted to the execution body described above.

After the description information is acquired, various methods can be flexibly adopted to determine the keywords according to the description information. For example, the description information may be processed using various existing keyword extraction methods to extract keywords.

Optionally, the description information may be processed by using a pre-trained keyword determination model to obtain a keyword corresponding to the description information. The keyword determination model can be used for representing the corresponding relation between the description information and the keywords.

It should be noted that the keyword determined according to the description information may be sub-information of the description information or may not be sub-information of the description information. For example, when the description information is a text, the keyword determined according to the description information may be a word included in the text, or may be a word obtained by performing adjustment such as retouching on a word included in the text.

The keyword determination model may be trained in advance based on existing machine learning methods. As an example, the keyword determination model may be trained by the following steps:

step one, training data is obtained.

In this step, the training data may include a set of description information and a keyword corresponding to each piece of description information in the set of description information. Specifically, a technician may specify an information set in advance, describe each piece of information in the information to obtain description information, and then extract a keyword of the description information by using a keyword extraction method or manually. The pre-specified information set can be specified according to actual application requirements. For example, if the target information desired to be generated is a video, the pre-specified information set may be a video set.

And step two, acquiring an initial keyword determination model, training the initial keyword determination model by using the acquired training data, and determining the initial keyword determination model after the training as a keyword determination model.

In this step, the initial keyword determination model may be various network models. For example, the initial keyword determination model may be a CNN (Convolutional Neural Networks), RNN (Recurrent Neural Networks), Transformer model, or the like. Specifically, the existing neural network model which is trained or not trained can be directly used as the initial keyword determination model, and technicians can design and utilize deep learning frameworks such as keras and Tensflow according to requirements to build the initial keyword determination model.

After the initial keyword determination model is obtained, description information can be selected from training data to serve as input of the initial keyword determination model, meanwhile, keywords corresponding to the selected description information in the training data serve as expected output of the initial keyword determination model, and parameters of each network layer of the initial keyword determination model are continuously adjusted by using algorithms such as gradient descent, back propagation and the like on the basis of a preset loss function, so that the initial keyword determination model is trained, and the trained keyword determination model is obtained. Wherein the loss function can be pre-designed by the skilled person according to the actual requirements.

For the target information desired to be generated, the content or effect, etc. desired to be presented by the designer of the target information can be conveyed more accurately due to the descriptive information. Therefore, the description information of the target information is obtained firstly, and the keyword is determined based on the description information to assist in generating the target information, so that the accuracy of the keyword can be improved to a certain extent, the matching degree of the generated target information and the actual expectation is improved, and the quality of the generated target information is improved.

In some optional implementation manners of this embodiment, after the keyword is obtained, the obtained keyword may be used as a search keyword, and a search is performed by using the keyword to obtain a search result. Wherein, the retrieval result can comprise at least one piece of information obtained by the retrieval by the keyword and the content label of each piece of information obtained by the retrieval. Then, a content tag corresponding to each candidate information set of the keyword and the candidate information set may be determined according to the search result.

Specifically, the keyword may be used for searching in a database or a search engine to obtain a search result. By way of example, when the target information is a video, the target information can be retrieved in a search engine by using keywords, so that a plurality of related videos can be retrieved. Currently, many retrieved videos carry tags themselves. For example, a video producer or uploader adds tags for videos. Also for example, a presentation platform of the video or a tab added by the viewer for the video, and so on. In this case, the tag carried by the retrieved video itself may be used as the content tag of the video.

After obtaining the search result, information may be selected from the at least one piece of information searched as candidate information, so as to obtain a candidate information set, and a content tag of each selected information may be obtained. And selecting information from the at least one piece of information obtained by retrieval as candidate information by adopting various methods according to requirements. For example, a preset number of pieces of information may be selected as candidate information in order from new to old in the production time of the information. For another example, a preset number of pieces of information may be randomly selected as the candidate information. For another example, a preset number of pieces of information may be selected from at least one piece of information as candidate information in an order from a large degree to a small degree of association between the information and the keyword.

Because various information displayed in the internet at present has content labels, the content labels of the directly acquired information can effectively improve the acquisition speed of the content labels. Meanwhile, the content tags are provided by the creator of the corresponding information or the viewer of the information, so that the content tags have high accuracy, and on the basis, the accuracy of the determined index value of the corresponding information can be improved by using the content tags, so that the quality of the screened material is improved, and the high-quality target information is generated.

In some optional implementation manners of the present embodiment, for each candidate information in the candidate information set, the content label of the candidate information and the obtained keyword may be processed by using a pre-trained index value prediction model to obtain an index value of the candidate information. The index value prediction model can be used for representing the corresponding relation between the content labels and the keywords of the candidate information and the index values of the candidate information.

The index value prediction model can be obtained by pre-training based on the existing machine learning method. By way of example, the index value prediction model may be trained by:

step one, training data is obtained.

In this step, the training data may include content tags and keywords corresponding to at least one piece of information, and also include index values corresponding to at least one piece of information. Specifically, at least one piece of information may be acquired in advance by a technician, and then a content tag and a keyword of each piece of information may be set while setting an index value of each piece of information.

For example, when the target information to be generated is a video, a technician may acquire at least one video in advance, then acquire a content tag of each video from the network, and manually set a keyword and an index value of each video.

And step two, acquiring an initial index value prediction model, then training the initial index value prediction model by using the acquired training data, and determining the trained initial index value prediction model as an index value prediction model.

In this step, the initial index value prediction model may be various network models. For example, the initial index value prediction model may be CNN (Convolutional Neural Networks), RNN (Recurrent Neural Networks), or the like. Specifically, the existing neural network model which is trained or not trained can be directly used as the initial index value prediction model, and technicians can design and utilize deep learning frameworks such as keras and Tensflow according to requirements to build the initial index value prediction model.

After the initial index value prediction model is obtained, the content labels and the keywords corresponding to the content labels can be selected from the training data to serve as the input of the initial index value prediction model, the index values corresponding to the selected content labels in the training data serve as the expected output of the initial index value prediction model, and the parameters of each network layer of the initial index value prediction model are continuously adjusted by using algorithms such as gradient descent, back propagation and the like based on a preset loss function, so that the initial index value prediction model is trained, and the trained index value prediction model is obtained. Wherein the loss function can be pre-designed by the skilled person according to the actual requirements.

The method has the advantages that training data are set according to actual requirements to train the index value prediction model, so that the index value of each piece of information can be quickly obtained by using the index value prediction model, and compared with the method of manually setting the index value of each piece of information, the method can improve the speed of determining the index value of the information, avoid subjectivity of different users to the index value of the information, and improve the accuracy of the determined index value of the information.

Alternatively, for each candidate information in the candidate information set, the content label of the candidate information, the acquired keyword, and the candidate information may be processed by using a pre-trained index value prediction model to obtain an index value of the candidate information. In this case, the pre-trained index value prediction model may be used to characterize the correspondence between the candidate information, the content label of the candidate information, the corresponding keyword, and the index value of the candidate information. Specifically, the candidate information, the content tag of the candidate information, and the corresponding keyword may be used as input of the index value prediction model to obtain the index value of the candidate information.

Therefore, in the process of determining the index value of the candidate information, more analysis on the content of the candidate information is added, so that the accuracy of the determined index value of the information is further improved.

In some optional implementations of this embodiment, the index value may be used to characterize at least one of the following: the association degree of the candidate information and the acquired keyword, the quality of the candidate information, and the association degree of the sub information included in the target information expected to be generated and the candidate information.

Specifically, the degree of association between the candidate information and the obtained keyword can be determined according to actual application requirements, and what characterizes the degree of association between the candidate information and the obtained keyword in what aspects. Similarly, it may be determined according to actual application requirements, to which aspect the quality of the candidate information specifically refers, and of course, it may also be determined according to actual application requirements, that the degree of association between the sub information included in the target information that is desired to be generated and the candidate information characterizes the degree of association between the sub information included in the target information that is desired to be generated and the candidate information in which aspect.

The sub information included in the target information desired to be generated may be various contents appearing in the target information. For example, when the object information is an image, the sub information included in the object information may be a certain image area of the image or may be something appearing in the image. For another example, when the target information is a video, the sub information included in the target information may be one of video clips of the video, or may be people and objects appearing in the video.

The information material expected to be screened out can be limited from multiple aspects through one or more of the association degree of the candidate information and the acquired keywords, the quality of the candidate information and the association degree of the sub information and the candidate information included in the target information expected to be generated, so that the information material which is more in line with the requirement, namely more matched with the target information expected to be generated is screened out, and the quality of the generated target information is guaranteed.

In some optional implementations of this embodiment, the target information may include information of the target person. At this time, the acquired keywords may include an identification of the event associated with the target person and an identification of the target person.

In this embodiment, the information of the target person may include information of various aspects of the target person, and may be determined according to actual needs. For example, the target information may be information for introducing a target person. As an example, when the target information is an image, the target information may be just an image of a target person. When the target information is audio, the target information may be introduction audio of the target person. When the target information is a video, the target information may be an introduction video of a target person.

The target person-associated events may include various events related or related to the target person. For example, the target person-associated event may include an event in which the target person participates. Taking the target information as the introduction video of the person X as an example, the events related to the person X may include various events occurring in the life of the person X. Such as the birth of character X, the acquisition of a prize, the shooting of a movie work or design work, and so on. The identification of the event may be a keyword, code number, etc. of the event. The identification of the target person may be the name, nickname, etc. of the target person.

When the target information relates to the target person, the identifier of the event related to the target person and the identifier of the target person can be set intentionally as keywords, so that information with high relevance of the target person can be screened out to generate related information for the target person. Particularly, under the application scenes of needing the information of the target person, the method can be used for greatly shortening the material selection time, so that the production period of the target information is shortened.

The method provided by the above embodiment of the present disclosure generates the target information by acquiring the keyword corresponding to the target information desired to be generated, then retrieving the corresponding information as the candidate information by using the keyword, then determining the index value of each candidate information by using the keyword and the content tag of each candidate information, and then selecting the candidate information as the information material of the target information according to the index value, thereby generating the target information by using the selected information material. Compared with the manual screening of information materials from massive information, the method can greatly reduce the screening time of the information materials, thereby improving the production efficiency of the target information.

With further reference to fig. 3, a flow 300 of yet another embodiment of a method for generating information is shown. The flow 300 of the method for generating information comprises the steps of:

step 301, obtaining description information corresponding to at least one target scene respectively.

In this embodiment, the target scene may be a scene expressed or represented by a target video desired to be generated. The description information of each target scene can be set by technicians such as a designer of the target information according to design requirements.

Step 302, for a target scene in at least one target scene, the following step 3021-3025 is performed to obtain a video corresponding to the target scene.

Step 3021, determining the keyword of the target scene according to the description information of the target scene.

Step 3022, a candidate information set corresponding to the keyword of the target scene is obtained, and a content tag of the candidate information in the candidate information set is obtained.

Wherein the candidate information set is a candidate video set and/or a candidate image set.

Step 3023, for the candidate information in the candidate information set, determining an index value of the candidate information according to the content tag and the keyword of the candidate information.

And step 3024, selecting the candidate information from the candidate information set as the information material corresponding to the target scene according to the index values corresponding to the candidate information in the candidate information set.

And step 3025, generating a video corresponding to the target scene according to the information material corresponding to the target scene.

The specific execution process of the steps 3021-3025 can refer to the related descriptions of the

steps

201 and 205 in the corresponding embodiment of fig. 2, and will not be described herein again.

Step 303, generating target videos according to the videos corresponding to the at least one target scene.

After the video corresponding to each target scene is obtained, the respective videos may be further adjusted and fused by using various existing video production methods to generate a target video desired to be generated.

With continued reference to fig. 4, fig. 4 is an illustrative application scenario 400 of the method for generating information according to the present embodiment. In the application scenario of fig. 4, description information of two scenes included in a video that a video producer desires to generate may be obtained first. As shown by

reference numerals

401 and 402 in the figure, a scene of one scene is described as "ancient decoration phase of character X", and a scene of the other scene is described as "ancient decoration phase of character Y".

The scene description 401 of the first scene may then be input to a pre-trained keyword determination model 403, resulting in keywords 404 of the scene description 401, including X and antiques. Likewise, a scene description 402 of another scene may be input to a pre-trained keyword determination model 403, resulting in keywords 405 of the scene description 402, including Y and antiques.

Then, a search may be performed in a preset database using the obtained keyword 404 to obtain a search result 406, where the search result 406 includes a plurality of videos and a content tag of each video that are searched using the keyword 404. For each retrieved video, the content tags and keywords 404 for that video may then be input to a pre-trained predictive model 408, resulting in a score for that video. And screening out videos with scores larger than a preset threshold value from the retrieved scores of the videos as a material 409, and making a video 411 corresponding to the scene description 401 by using the material 409.

Correspondingly, a search result 407 can be obtained by searching in a preset database by using the obtained keyword 405, in this case, the search result 407 includes a plurality of videos searched by using the keyword 405 and a content tag of each video. For each retrieved video, the content tags and keywords 405 for that video may then be input to a pre-trained predictive model 408, resulting in a score for that video. And screening out videos with scores larger than a preset threshold value from the retrieved scores of the videos as a material 410, and making a video 412 corresponding to the scene description 402 by using the material 410.

Finally, a video 413 for presenting the scene description 401 and the scene description 402 may be synthesized from the generated video 411 for presenting the scene description 401 and the video 412 for presenting the scene description 402. This makes it possible to obtain a mixed cut video 413 of the ancient fashion looks of the character X and the character Y.

As can be seen from fig. 3, compared with the embodiment corresponding to fig. 2, the method for generating information in this embodiment may split a video that is expected to be generated according to a certain granularity, determine a keyword for each scene and screen a corresponding information material, so as to generate a video corresponding to each scene to present the scene, and then synthesize the video according to the videos corresponding to the scenes, thereby ensuring accuracy of the screened material for each scene, further improving quality of the synthesized video, and ensuring detail accuracy of the synthesized video. In addition, due to the splitting of the scenes, all the scenes can be processed simultaneously, so that the video generation efficiency is further improved, and the video generation period is greatly shortened.

With further reference to fig. 5, as an implementation of the methods shown in the above figures, the present disclosure provides an embodiment of an apparatus for generating information, which corresponds to the method embodiment shown in fig. 2, and which is particularly applicable in various electronic devices.

As shown in fig. 5, the apparatus 500 for generating information provided by the present embodiment includes a first acquiring unit 501, a second acquiring unit 502, a determining unit 503, a selecting unit 504, and a generating unit 505. Wherein the first obtaining unit 501 is configured to obtain a keyword; the second acquisition unit 502 is configured to acquire a candidate information set corresponding to the keyword, and acquire content tags of candidate information in the candidate information set; the determination unit 503 is configured to determine, for candidate information in the candidate information set, an index value of the candidate information according to the content tag and the keyword of the candidate information; the selecting unit 504 is configured to select candidate information from the candidate information set as information material according to the index value corresponding to each candidate information in the candidate information set; the generating unit 505 is configured to generate target information from the information material.

In the present embodiment, in the apparatus 500 for generating information: the detailed processing and the technical effects of the first obtaining unit 501, the second obtaining unit 502, the determining unit 503, the selecting unit 504 and the generating unit 505 can refer to the related descriptions of step 201, step 202, step 203, step 204 and step 205 in the corresponding embodiment of fig. 2, which are not described herein again.

In the apparatus provided by the above embodiment of the present disclosure, the first obtaining unit obtains the keyword; the second acquisition unit acquires a candidate information set corresponding to the keyword and acquires a content tag of candidate information in the candidate information set; the determining unit determines an index value of the candidate information according to the content tag and the keyword of the candidate information for the candidate information in the candidate information set; the selection unit selects the candidate information from the candidate information set as an information material according to the index values corresponding to the candidate information in the candidate information set respectively; the generation unit generates target information from the information material. Therefore, the information material of the target information expected to be generated can be quickly acquired from the mass information, and the generation period of the target information is reduced.

Referring now to FIG. 6, a schematic diagram of an electronic device (e.g., the server of FIG. 1) 600 suitable for use in implementing embodiments of the present disclosure is shown. The terminal device in the embodiments of the present disclosure may include, but is not limited to, a mobile terminal such as a mobile phone, a notebook computer, a digital broadcast receiver, a PDA (personal digital assistant), a PAD (tablet computer), a PMP (portable multimedia player), a vehicle terminal (e.g., a car navigation terminal), and the like, and a fixed terminal such as a digital TV, a desktop computer, and the like. The server shown in fig. 6 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present disclosure.

As shown in fig. 6, electronic device 600 may include a processing means (e.g., central processing unit, graphics processor, etc.) 601 that may perform various appropriate actions and processes in accordance with a program stored in a Read Only Memory (ROM)602 or a program loaded from a storage means 608 into a Random Access Memory (RAM) 603. In the RAM603, various programs and data necessary for the operation of the electronic apparatus 600 are also stored. The processing device 601, the ROM 602, and the RAM603 are connected to each other via a bus 604. An input/output (I/O) interface 605 is also connected to bus 604.

Generally, the following devices may be connected to the I/O interface 605: input devices 606 including, for example, a touch screen, touch pad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; output devices 607 including, for example, a Liquid Crystal Display (LCD), a speaker, a vibrator, and the like; storage 608 including, for example, tape, hard disk, etc.; and a communication device 609. The communication means 609 may allow the electronic device 600 to communicate with other devices wirelessly or by wire to exchange data. While fig. 6 illustrates an electronic device 600 having various means, it is to be understood that not all illustrated means are required to be implemented or provided. More or fewer devices may alternatively be implemented or provided. Each block shown in fig. 6 may represent one device or may represent multiple devices as desired.

In particular, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network via the communication means 609, or may be installed from the storage means 608, or may be installed from the ROM 602. The computer program, when executed by the processing device 601, performs the above-described functions defined in the methods of embodiments of the present disclosure.

It should be noted that the computer readable medium described in the embodiments of the present disclosure may be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, 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 disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In embodiments of the disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In embodiments of the present disclosure, however, a computer readable signal medium may comprise a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: electrical wires, optical cables, RF (radio frequency), etc., or any suitable combination of the foregoing.

Computer program code for carrying out operations for embodiments of the present disclosure may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + +, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

In accordance with one or more embodiments of the present disclosure, there is provided a method for generating information, the method comprising: acquiring a keyword; acquiring a candidate information set corresponding to the keyword, and acquiring a content tag of candidate information in the candidate information set; for candidate information in the candidate information set, determining an index value of the candidate information according to the content tag and the keyword of the candidate information; selecting candidate information from the candidate information set as an information material according to index values corresponding to the candidate information in the candidate information set respectively; and generating target information according to the information material.

According to one or more embodiments of the present disclosure, determining an index value of the candidate information according to the content tag and the keyword of the candidate information includes: and processing the content labels and the keywords of the candidate information by using a pre-trained index value prediction model to obtain the index value of the candidate information.

According to one or more embodiments of the present disclosure, acquiring a candidate information set corresponding to a keyword, and acquiring a content tag of candidate information in the candidate information set, includes: searching by using the keyword to obtain a search result, wherein the search result comprises at least one piece of information searched according to the keyword and a content label of the information in the at least one piece of information; and determining the candidate information set and the content label of the candidate information in the candidate information set according to the retrieval result.

According to one or more embodiments of the present disclosure, obtaining a keyword includes: acquiring description information of target information; and determining the keywords according to the description information.

According to one or more embodiments of the present disclosure, determining a keyword according to the description information includes: and processing the description information by utilizing a pre-trained keyword determination model to obtain keywords corresponding to the description information.

According to one or more embodiments of the present disclosure, the indicator value is used to characterize at least one of the following indicators: the association degree of the candidate information and the keyword, the quality of the candidate information, and the association degree of the sub information included in the target information and the candidate information.

According to one or more embodiments of the present disclosure, the target information includes information of a target person; and the keywords include an identification of the event associated with the target person and an identification of the target person.

In accordance with one or more embodiments of the present disclosure, there is provided an apparatus for generating information, the apparatus comprising: a first acquisition unit configured to acquire a keyword; a second acquisition unit configured to acquire a candidate information set corresponding to the keyword, and acquire a content tag of candidate information in the candidate information set; a determination unit configured to determine, for candidate information in the candidate information set, an index value of the candidate information according to a content tag and a keyword of the candidate information; the selection unit is configured to select the candidate information from the candidate information set as an information material according to the index values respectively corresponding to the candidate information in the candidate information set; a generating unit configured to generate target information from the information material.

According to one or more embodiments of the present disclosure, the determining unit is further configured to process the content tags and keywords of the candidate information by using a pre-trained index value prediction model, so as to obtain the index value of the candidate information.

According to one or more embodiments of the present disclosure, the second obtaining unit is further configured to perform a search by using the keyword, and obtain a search result, where the search result includes at least one piece of information searched according to the keyword and a content tag of information in the at least one piece of information; and determining the candidate information set and the content label of the candidate information in the candidate information set according to the retrieval result.

According to one or more embodiments of the present disclosure, the first obtaining unit is further configured to obtain description information of the target information; and determining the keywords according to the description information.

According to one or more embodiments of the present disclosure, the first obtaining unit is further configured to process the description information by using a pre-trained keyword determination model, and obtain a keyword corresponding to the description information.

According to one or more embodiments of the present disclosure, the index value is used for characterizing at least one of the following indexes: the association degree of the candidate information and the keyword, the quality of the candidate information, and the association degree of the sub information included in the target information and the candidate information.

The units described in the embodiments of the present disclosure may be implemented by software or hardware. The described units may also be provided in a processor, and may be described as: a processor includes a first obtaining unit, a second obtaining unit, a determining unit, a selecting unit, and a generating unit. Where the names of these units do not in some cases constitute a limitation on the unit itself, for example, the first acquisition unit may also be described as a "unit that acquires a keyword".

As another aspect, the present disclosure also provides a computer-readable medium. The computer readable medium may be embodied in the electronic device described above; or may exist separately without being assembled into the electronic device. The computer readable medium carries one or more programs which, when executed by the electronic device, cause the electronic device to: acquiring a keyword; acquiring a candidate information set corresponding to the keyword, and acquiring a content tag of candidate information in the candidate information set; for candidate information in the candidate information set, determining an index value of the candidate information according to the content tag and the keyword of the candidate information; selecting candidate information from the candidate information set as an information material according to index values corresponding to the candidate information in the candidate information set respectively; and generating target information according to the information material.

The foregoing description is only exemplary of the preferred embodiments of the disclosure and is illustrative of the principles of the technology employed. It will be appreciated by those skilled in the art that the scope of the invention in the embodiments of the present disclosure is not limited to the specific combination of the above-mentioned features, but also encompasses other embodiments in which any combination of the above-mentioned features or their equivalents is made without departing from the inventive concept as defined above. For example, the above features and (but not limited to) technical features with similar functions disclosed in the embodiments of the present disclosure are mutually replaced to form the technical solution.

Claims

1. A method for generating information, comprising:

acquiring a keyword;

acquiring a candidate information set corresponding to the keyword, and acquiring a content tag of candidate information in the candidate information set;

for candidate information in the candidate information set, determining an index value of the candidate information according to the content tag of the candidate information and the keyword;

selecting candidate information from the candidate information set as an information material according to index values corresponding to the candidate information in the candidate information set respectively;

and generating target information according to the information material.

2. The method of claim 1, wherein determining the index value of the candidate information according to the content tag of the candidate information and the keyword comprises:

and processing the content label of the candidate information and the keyword by using a pre-trained index value prediction model to obtain the index value of the candidate information.

3. The method of claim 1, wherein the obtaining a candidate information set corresponding to the keyword and obtaining content tags of candidate information in the candidate information set comprises:

searching by using the keyword to obtain a search result, wherein the search result comprises at least one piece of information searched according to the keyword and a content label of the information in the at least one piece of information;

and determining the candidate information set and the content label of the candidate information in the candidate information set according to the retrieval result.

4. The method of claim 2, wherein the obtaining keywords comprises:

acquiring description information of the target information;

and determining the keywords according to the description information.

5. The method of claim 4, wherein the determining the keyword from the description information comprises:

and processing the description information by utilizing a pre-trained keyword determination model to obtain a keyword corresponding to the description information.

6. The method according to one of claims 1 to 5, wherein the indicator value is used to characterize at least one of the following indicators: the association degree of the candidate information and the keyword, the quality of the candidate information, and the association degree of the sub information included in the target information and the candidate information.

7. The method of any of claims 1-5, wherein the target information includes information of a target person; and

the keywords include an identification of an event associated with the target person and an identification of the target person.

8. An apparatus for generating information, wherein the apparatus comprises:

a first acquisition unit configured to acquire a keyword;

a second acquisition unit configured to acquire a candidate information set corresponding to the keyword, and acquire a content tag of candidate information in the candidate information set;

a determination unit configured to determine, for candidate information in the candidate information set, an index value of the candidate information according to a content tag of the candidate information and the keyword;

the selecting unit is configured to select candidate information from the candidate information set as information materials according to index values corresponding to the candidate information in the candidate information set respectively;

a generating unit configured to generate target information from the information material.

9. An electronic device, comprising:

one or more processors;

a storage device having one or more programs stored thereon;

when executed by the one or more processors, cause the one or more processors to implement the method of any one of claims 1-7.

10. A computer-readable medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1-7.