Content item processing method, apparatus, device, and storage medium

Abstract

The application discloses a content item processing method, device and equipment and a storage medium, and belongs to the technical field of computers. The method comprises the following steps: performing identification processing on retrieval data to obtain retrieval intention information of the retrieval data, wherein the retrieval intention information is used to represent the retrieval intention of the retrieval data; generating a content item matched with the retrieval data based on the retrieval intention information; and putting the content item based on the retrieval data. According to the scheme, the retrieval intention information of the retrieval data is obtained, the suitable content item is generated, and the content item is put based on the retrieval data, so that the retrieval data and the content item are matched, the situation that the content item is not put in the matched retrieval data is avoided, and the putting effect is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular to a content item processing method and device, equipment and storage medium. BACKGROUND

[0002] In the SEM (Search Engine Marketing, search engine marketing) scene, the content item such as an advertisement corresponds to a delivery word, and when the search data input by the user matches the delivery word of the content item, the content item is presented to the user as a search result.

[0003] In the related art, according to the business scope of the advertiser, a search object (for example, the business scope is tourism, and the search object is scenic spot A) matching the business scope is determined, the search data related to the search object is manually sorted, and a high-frequency template (the business hours of xxx) of the search data is obtained. According to the high-frequency template, the text content of the content item is generated, the image content of the content item is obtained according to the search object, the text content and the image content are combined into the content item, and the search object and the high-frequency template are combined into the delivery word (the business hours of scenic spot A) of the content item.

[0004] However, it is found in actual application that when the search data input by the user is "scenic spot A opens from how many o'clock to how many o'clock" and the like, which is quite different from the expression of the delivery word, it is determined that the search data does not match the delivery word, so the content item will not be delivered, resulting in poor delivery effect of the content item. SUMMARY

[0005] Embodiments of the present application provide a content item processing method, device, equipment and storage medium, which improves the delivery effect of the content item. The technical solution is as follows:

[0006] On the one hand, a content item processing method is provided, which comprises:

[0007] identifying the search data to obtain search intent information of the search data, the search intent information being used to represent the search intent of the search data;

[0008] generating a content item matching the search data based on the search intent information;

[0009] delivering the content item based on the search data.

[0010] On the one hand, a content item processing device is provided, which comprises:

[0011] an identification module, configured to identify the search data to obtain search intent information of the search data, the search intent information being used to represent the search intent of the search data;

[0012] generating, based on the search intention information, a content item matching the search data;

[0013] delivering, based on the search data, the content item.

[0014] In a possible implementation, the identification module is configured to perform at least one of the following:

[0015] The search intention information includes an intention category corresponding to the search data, the intention category being used to represent a category to which a search intention of the search data belongs; the search data is subjected to first identification processing to obtain the intention category corresponding to the search data.

[0016] The search intention information includes a business category corresponding to the search data, the business category representing a business category to which a search object of the search data belongs; the search data is subjected to second identification processing to obtain the business category corresponding to the search data.

[0017] The search intention information includes a search object of the search data; the search data is subjected to third identification processing to obtain the search object of the search data.

[0018] In a possible implementation, the identification module is configured to perform first identification processing on the search data based on a business category corresponding to the search data, and determine, from a plurality of intention categories corresponding to the business category, an intention category corresponding to the search data.

[0019] In a possible implementation, the identification module includes:

[0020] An acquisition unit is configured to acquire an intention identification model, the intention identification model including a feature extraction layer, a feature processing layer, and an identification layer corresponding to each business category, the identification layer corresponding to the business category being used to determine, from a plurality of intention categories corresponding to the business category, an intention category to which a search intention of search data belongs.

[0021] A first processing unit is configured to perform feature extraction on the search data through the feature extraction layer to obtain first feature data.

[0022] A second processing unit is configured to perform processing on the first feature data through the feature extraction layer to obtain second feature data.

[0023] An input unit is configured to input, based on a business category corresponding to the search data, the second feature data to the identification layer corresponding to the business category.

[0024] The third processing unit is configured to perform identification processing on the second feature data through the identification layer to obtain an intent category corresponding to the search data.

[0025] In a possible implementation, the first processing unit is configured to perform word segmentation processing on the search data to obtain a plurality of words;

[0026] The first processing unit is configured to perform feature extraction on each word in the plurality of words through the feature extraction layer to obtain first feature data corresponding to the word.

[0027] The second processing unit is configured to, for the first feature data corresponding to each word, perform fusion processing on the first feature data corresponding to the word and first feature data of other words in the plurality of words except the word through the feature extraction layer to obtain second feature data of the word.

[0028] In a possible implementation, the apparatus further includes a training module.

[0029] The training module is configured to obtain a plurality of sample data, where the plurality of sample data includes a plurality of sample search data corresponding to different service categories and sample intent categories corresponding to the plurality of sample search data respectively.

[0030] The training module is configured to train an initial intent identification model based on the plurality of sample data to obtain an intent identification model including a target identification layer, where the target identification layer is an identification layer corresponding to a plurality of service categories.

[0031] The training module is configured to perform copy processing on the target identification layer according to a number of service categories to obtain an identification layer corresponding to each service category.

[0032] The training module is configured to, for the identification layer corresponding to each service category, train the identification layer corresponding to the service category based on a plurality of sample search data corresponding to the category and sample intent categories corresponding to the plurality of sample search data.

[0033] In a possible implementation, the generation module is configured to perform at least one of the following:

[0034] Based on the intent category, obtain text data corresponding to the intent category from a first correspondence relationship, where the first correspondence relationship is a correspondence relationship between an intent category and text data, and the text data is text data describing a content item; and based on the text data, generate a content item matching the search data.

[0035] based on the image data, generate a content item matching the search data;

[0036] in a case where the business category corresponding to the search data is a target business category, generate, based on the search intent information, a content item matching the search data, the target business category being a business category to which the content item to be recommended belongs;

[0037] based on the search object, obtain a link of the search object; and based on the link of the search object, generate a content item matching the search data.

[0038] In a possible implementation, the search data is search data obtained in response to a search operation, and the delivery module is configured to display the content item as a search result corresponding to the search data.

[0039] In a possible implementation, the search data is search data obtained from a search record, and the delivery module includes:

[0040] a determination unit configured to determine the search data as delivery data of the content item;

[0041] an obtaining unit configured to obtain input search data in response to a search operation;

[0042] a display unit configured to, in a case where the input search data matches the delivery data, display the content item when displaying a search result of the search data.

[0043] In a possible implementation, the identification module is configured to perform identification processing on a plurality of pieces of search data to obtain search intent information of each piece of search data; and based on the search intent information of each piece of search data, determine at least one piece of search data corresponding to different search intent information.

[0044] The generation module is configured to determine the at least one piece of search data corresponding to the search intent information as delivery data of a content item matching the search intent information.

[0045] In an aspect, a computer device is provided, which includes one or more processors and one or more memories having stored therein at least one program code, which is loaded and executed by the one or more processors to implement operations performed by the content item processing method according to any of the possible implementations.

[0046] In an aspect, a computer-readable storage medium is provided, the storage medium having stored therein at least one program code, the at least one program code being loadable by a processor and executable by the processor to implement operations performed by the content item processing method according to any of the possible implementation manners described above.

[0047] In an aspect, a computer program or computer program product is provided, the computer program or computer program product comprising: computer program code which, when executed by a computer, causes the computer to implement operations performed by the content item processing method according to any of the possible implementation manners described above.

[0048] The content item processing method, apparatus, device and storage medium provided by the embodiments of the present application can obtain the search intention information of search data, generate suitable content items, and perform delivery of the content items based on the search data, so as to ensure matching between the search data and the content items, avoid the situation that no content item is delivered in the matching search data, and improve the delivery effect. BRIEF DESCRIPTION OF DRAWINGS

[0049] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0050] Figure 1 is a schematic diagram of an implementation environment provided by the embodiments of the present application;

[0051] Figure 2 is a flowchart of a content item processing method provided by the embodiments of the present application;

[0052] Figure 3 is a flowchart of a content item processing method provided by the embodiments of the present application;

[0053] Figure 4 is a schematic diagram of an object recognition model provided by the embodiments of the present application;

[0054] Figure 5 is a schematic diagram of an intention recognition model provided by the embodiments of the present application;

[0055] Figure 6 is a flowchart of a content item processing method provided by the embodiments of the present application;

[0056] Figure 7 is a flowchart of a content item processing method provided by the embodiments of the present application;

[0057] Figure 8is a flowchart of a data retrieval processing method provided by an embodiment of the present application;

[0058] Figure 9 is a structural schematic diagram of a content item processing apparatus provided by an embodiment of the present application;

[0059] Figure 10 is another structural schematic diagram of a content item processing apparatus provided by an embodiment of the present application;

[0060] Figure 11 is a structural schematic diagram of a terminal provided by an embodiment of the present application;

[0061] Figure 12 is a structural schematic diagram of a server provided by an embodiment of the present application. DETAILED DESCRIPTION

[0062] To make the objectives, technical solutions and advantages of the present application clearer, the following will further describe the embodiments of the present application in detail with reference to the accompanying drawings.

[0063] It can be understood that the terms "first", "second" and the like as used in the present application can be used in the description of various concepts in this document, but unless specifically stated, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the present application, a first image can be referred to as a second image, and similarly, a second image can be referred to as a first image.

[0064] The terms "at least one", "multiple", "each", "any" used in the present application include one, two or more than two, multiple includes two or more than two, and each refers to each of the corresponding multiple, any refers to any one of the multiple, for example, multiple images include 3 images, and each refers to each of the 3 images, any refers to any one of the 3 images, which can be the first, the second or the third.

[0065] The content item processing method provided by the embodiments of the present application is executed by a computer device. In a possible implementation manner, the computer device is a terminal, for example, the terminal is any type of terminal such as a desktop computer, a tablet computer or a mobile phone. In another possible implementation manner, the computer device is a server. For example, the server can be a server, or a server cluster composed of several servers, or a cloud computing service center. In another possible implementation manner, the computer device includes a terminal and a server.

[0066] Figure 1 is a schematic diagram of an implementation environment provided by an embodiment of the present application, as Figure 1As shown, the implementation environment includes a terminal 101 and a server 102. The terminal 101 and the server 102 are connected through a wireless or wired network.

[0067] Optionally, the terminal 101 is any type of terminal such as a desktop computer, a tablet computer or a mobile phone. The server 102 is a server, a server cluster composed of a plurality of servers or a cloud computing service center.

[0068] The terminal 101 installs a target application provided by the server 102, and the terminal 101 can implement functions such as data transmission and message interaction through the target application. Optionally, the target application is an application in the operating system of the terminal 101 or an application provided by a third party. For example, the target application is a multimedia data processing application that has a function of processing multimedia data, and of course, the multimedia data processing application can also have other functions such as a comment function and a sharing function.

[0069] Optionally, the terminal 101 obtains search data, sends the search data to the server 102, and the server 102 performs identification processing on the search data to obtain search intent information of the search data, the search intent information being used to represent a search intent of the search data; the server 102 generates a content item matched with the search data based on the search intent information, and delivers the content item based on the search data. For example, the server 102 displays the content item as a search result corresponding to the search data; for another example, the server 102 takes the search data as delivery data of the content item, and next time after obtaining input search data, if the input search data matches the delivery data, the content item is displayed when the search result of the search data is displayed.

[0070] The content item processing method provided by the embodiments of the present application can be applied to any content item delivery scenario, and the embodiments of the present application do not limit the content item delivery scenario.

[0071] Figure 2 FIG. 1 is a flowchart of a content item processing method provided by an embodiment of the present application. The embodiment of the present application takes a computer device as an example of an execution subject, and the embodiment includes the following steps.

[0072] 201. The computer device performs identification processing on search data to obtain search intent information of the search data, the search intent information being used to represent a search intent of the search data.

[0073] The retrieval data is data input by a user for retrieval purposes. In the embodiment of the present application, the retrieval data is subjected to identification processing to identify the retrieval intention of the retrieval data and obtain retrieval intention information of the retrieval data. The retrieval intention information is used to indicate the retrieval intention of the retrieval data, which indicates what information the user wants to retrieve by inputting the retrieval information. For example, the retrieval data "opening hours of scenic spot A from time to time" has a retrieval intention of "opening hours of scenic spot A".

[0074] 202. The computer device generates a content item matching the retrieval data based on the retrieval intention information.

[0075] In the embodiment of the present application, after obtaining the retrieval intention information of the user, a content item matching the retrieval data can be generated based on the retrieval intention information, that is, a content item meeting the retrieval intention is generated according to the retrieval intention information of the user, and the content item is the content item matching the retrieval data.

[0076] 203. The computer device delivers the content item based on the retrieval data.

[0077] After generating the content item, the computer device can deliver the content item based on the retrieval data. Since the content item is generated based on the retrieval intention information of the retrieval data, the retrieval data meets the retrieval intention of the user, and delivering the content item to the user can meet the retrieval intention of the user, and also avoids the situation that no content item is delivered in the matching retrieval data, thereby improving the delivery effect of the content item.

[0078] The content item processing method provided in the embodiment of the present application can generate a suitable content item by obtaining retrieval intention information of retrieval data, and deliver the content item based on the retrieval data, which not only ensures that the retrieval data matches the content item, but also avoids the situation that no content item is delivered in the matching retrieval data, thereby improving the delivery effect.

[0079] It should be noted that the content item processing method provided in the embodiment of the present application can be applied to an online scenario or an offline scenario. When the embodiment of the present application is applied to an online scenario, the computer device delivers the content item based on the retrieval data, including: displaying the content item as a retrieval result corresponding to the retrieval data. When the embodiment of the present application is applied to an offline scenario, the computer device delivers the content item based on the retrieval data, including: taking the retrieval data as delivery data of the content item; in response to a retrieval operation, obtaining input retrieval data; in the case that the input retrieval data matches the delivery data, displaying the content item when displaying the retrieval result of the retrieval data. The embodiment of the present application takes an online scenario as an example for illustrative description. Figure 3 ​

[0080] Figure 3 is a flowchart of a content item processing method provided by an embodiment of the present application. An embodiment of the present application takes a computer device as an example for exemplary description, and the embodiment includes the following steps.

[0081] 301, the computer device acquires search data.

[0082] The search data is any data input by a user for searching, for example, data input by the user in a search box, and the like. Embodiments of the present application do not limit the search data, and do not limit the content of the search data. For example, the content of the search data is "how much is the ticket for scenic spot A", "what interesting places are there in city A", and the like.

[0083] 302, the computer device performs identification processing on the search data to obtain search intention information of the search data, the search intention information being used to represent a search intention of the search data.

[0084] Embodiments of the present application consider that different users have different description habits, and the input search data is very diverse. In order to provide accurate content items for users in the face of different search data, embodiments of the present application will first identify the search data input by the user to obtain search intention information of the search data, and provide accurate content items for the user through the search intention information.

[0085] In embodiments of the present application, the search intention information can include at least one of an intention category corresponding to the search data, a business category corresponding to the search data, and a search object of the search data. Embodiments of the present application do not limit the search intention information.

[0086] In embodiments of the present application, the search intention of the user is divided into different categories. The intention category can be: one-day tour strategy, self-driving tour strategy, free scenic spot, scenic spot opening time, and the like. Embodiments of the present application do not limit the intention category, and the intention category can be divided according to actual needs.

[0087] In embodiments of the present application, the business category corresponding to the search data is used to represent a category of a business to which a content item corresponding to the search data belongs. The business category can be tourism, food, hotel, home decoration, and the like. The business category can be divided according to actual needs, and embodiments of the present application do not limit the business category.

[0088] In embodiments of the present application, the search object of the search data is used to represent an entity object to which the search data is directed. Optionally, the search object can be an entity component in the search data. For example, the search data is "how much is the ticket for XX park", and the entity component is "XX park". Therefore, the search object is "XX park".

[0089] In a possible implementation, the computer device performs identification processing on the search data to obtain search intention information of the search data, including at least one of the following:

[0090] (1) The search intention information includes an intention category corresponding to the search data, and the intention category is used to represent a category to which a search intention of the search data belongs; the computer device performs first identification processing on the search data to obtain the intention category corresponding to the search data.

[0091] It should be noted that the embodiments of the present application can use any identification method to perform first identification processing on the search data to obtain the intention category corresponding to the search data. The embodiments of the present application do not limit the identification method, but only exemplarily illustrate the following two ways.

[0092] Optionally, the computer device determines the intention category corresponding to the search data according to a first keyword, which can be obtained empirically, divided according to actual needs, divided by a technical person, defaulted by the computer device, or the like. The embodiments of the present application do not limit the first keyword. The computer device performs first identification processing on the search data to obtain the intention category corresponding to the search data, including: the computer device obtains a corresponding relationship between the first keyword and the intention category, performs matching processing on the search data based on the first keyword in the corresponding relationship, obtains the first keyword matched with the search data, obtains the intention category corresponding to the first keyword from the corresponding relationship, and determines the intention category corresponding to the first keyword as the intention category corresponding to the search data.

[0093] Optionally, the computer device performs first identification processing through an intention recognition model. The computer device performs first identification processing on the search data to obtain the intention category corresponding to the search data, including: the computer device performs first identification processing on the search data through the intention recognition model to obtain the intention category corresponding to the search data. The intention recognition model is a model used to identify the intention category of the search data. Optionally, the intention recognition model is used to determine the intention category corresponding to the search data from a plurality of intention categories. For example, the intention recognition model is used to determine a probability that the search data belongs to each intention category, and determines the intention category corresponding to the maximum probability as the intention category corresponding to the search data.

[0094] (2) The search intention information includes a business category corresponding to the search data, and the business category represents a business category to which a search object of the search data belongs; the computer device performs second identification processing on the search data to obtain the business category corresponding to the search data.

[0095] It should be noted that the embodiments of the present application can use any identification method to perform second identification processing on the search data to obtain the business category corresponding to the search data. The embodiments of the present application do not limit the identification method, but only exemplarily illustrate the following two ways.

[0096] Optionally, the computer device determines the business category corresponding to the search data based on the second keyword. This second keyword can be derived from experience, classified according to actual needs, classified by technical personnel, or defaulted to by the computer device. This embodiment of the application does not limit the scope of the second keyword. The computer device performs a second identification process on the search data to obtain the business category corresponding to the search data, including: obtaining the correspondence between the second keyword and the business category; performing matching processing on the search data based on the second keyword in the correspondence to obtain the second keyword that matches the search data; obtaining the business category corresponding to the second keyword from the correspondence; and determining the business category corresponding to the second keyword as the business category corresponding to the search data.

[0097] Optionally, the computer device performs a second identification process using a business identification model. The computer device performs this second identification process on the retrieved data to obtain the business category corresponding to the retrieved data, including: the computer device performs the second identification process on the retrieved data using a business identification model to obtain the business category corresponding to the retrieved data. This business identification model is a model used to identify the business category of the retrieved data. Optionally, the business identification model is used to determine the business category corresponding to the retrieved data from multiple business categories. For example, the business identification model is used to determine the probability that the retrieved data belongs to each business category, and the business category with the highest probability is determined as the business category corresponding to the retrieved data.

[0098] (3) The search intent information includes the search object of the search data; the search data is subjected to third identification processing to obtain the search object of the search data.

[0099] Optionally, the computer device performs a third identification process on the search data to obtain the search object of the search data, including: the computer device obtains a reference search object, performs matching processing between the search data and the reference search object to obtain a reference search object that matches the search data, and determines the reference search object as the search object of the search data.

[0100] Optionally, the computer device performs third recognition processing using an object recognition model. Optionally, the computer device performs third recognition processing on the retrieved data to obtain the retrieval object of the retrieved data, including: the computer device performs character segmentation processing on the retrieved data, inputs the segmented retrieved data into an object recognition model, performs third recognition processing on the segmented retrieved data through the object recognition model to obtain the component category of each character, and determines the retrieval object of the retrieved data based on the component category of each character.

[0101] For example, such as Figure 4As shown, the part corresponding to the target ingredient category in the search data is taken as the search object. For example, the search data is "how much is red-braised carp", and the object recognition model determines that the ingredient of "red-braised carp" is "delicacy", and takes "red-braised carp" corresponding to "delicacy" as the search object of the search data.

[0102] It should be noted that in the embodiments of the present application, the object recognition model can be a bert model or other speech processing model, and the embodiments of the present application do not limit the object recognition model.

[0103] It should be noted that when the computer device performs identification processing on the search data to obtain the search intent information of the search data, any one or more of the above steps can be performed, and the embodiments of the present application do not limit this. Furthermore, when the above one or more steps are performed, the embodiments of the present application can be performed synchronously or in a certain order, and the embodiments of the present application do not limit this.

[0104] In a possible implementation, different intent categories are divided for different business categories. For example, the business category is a tourism category, and the intent categories under the tourism category include 54 categories, such as "one-day tour strategy", "self-driving tour strategy", "free scenic spot", "family travel", and "tour route". The computer device first determines the business category corresponding to the search data, and then determines the intent category corresponding to the search data. Optionally, the computer device performs first identification processing on the search data to obtain the intent category corresponding to the search data, including: based on the business category corresponding to the search data, performing first identification processing on the search data, and determining the intent category corresponding to the search data from the multiple intent categories corresponding to the business category.

[0105] The first identification processing on the search data can be performed by a keyword matching method or an intent recognition model, and the embodiments of the present application do not limit this, but only exemplarily illustrate the first identification processing by the intent recognition model.

[0106] Optionally, the computer device performs first identification processing on the search data based on the business category corresponding to the search data, and determines the intent category corresponding to the search data from the multiple intent categories corresponding to the business category, including: the computer device acquires the intent recognition model corresponding to the business category based on the business category corresponding to the search data, performs first identification processing on the search data by the intent recognition model, and obtains the intent category of the search data, and the intent recognition model corresponding to any business category is used to determine the intent category corresponding to the search data from the multiple intent categories corresponding to the business category.

[0107] Optionally, the computer device performs first identification processing on the search data based on the business category corresponding to the search data, determines an intent category corresponding to the search data from a plurality of intent categories corresponding to the business category, including: obtaining an intent identification model, the intent identification model including a feature extraction layer, a feature processing layer, and an identification layer corresponding to each business category, the identification layer corresponding to the business category being configured to determine an intent category to which a search intent of the search data belongs from a plurality of intent categories corresponding to the business category; performing feature extraction on the search data through the feature extraction layer to obtain first feature data; processing the first feature data through the feature extraction layer to obtain second feature data; inputting the second feature data into the identification layer corresponding to the business category based on the business category corresponding to the search data; and performing identification processing on the second feature data through the identification layer to obtain the intent category corresponding to the search data.

[0108] In the method, the computer device performs feature extraction on the search data through the feature extraction layer to obtain first feature data, including: performing word segmentation processing on the search data to obtain a plurality of words; and performing feature extraction on each word in the plurality of words through the feature extraction layer to obtain first feature data corresponding to the each word.

[0109] In the method, the computer device processes the first feature data through the feature extraction layer to obtain second feature data, including: for the first feature data corresponding to the each word, performing fusion processing on the first feature data corresponding to the word and first feature data of other words in the plurality of words except the word through the feature extraction layer to obtain second feature data of the word.

[0110] The embodiments of the present application also provide a training method of an intent identification model, wherein each identification layer of the intent identification model is trained through sample data corresponding to different business categories. In a possible implementation manner, the intent identification model includes a plurality of identification layers, the number of the plurality of identification layers being the same as the number of the business category types, and the plurality of identification layers being in one-to-one correspondence. Before the computer device performs first identification processing on the search data based on the business category through the intent identification model, and determines an intent category corresponding to the search data from a plurality of intent categories corresponding to the business category, the method further includes: for each business category, obtaining a plurality of sample data corresponding to the business category, the sample data including sample search data and a sample intent category corresponding to the sample search data; and training a feature extraction layer in the intent identification model and an identification layer corresponding to the business category based on the plurality of sample data corresponding to each business category.

[0111] To improve the training efficiency of the intent recognition model, this application also provides another training method for the intent recognition model. First, a recognition layer is trained based on sample data from all business categories. Then, this recognition layer is copied to obtain a recognition layer corresponding to each business category. Each recognition layer is then trained specifically. In one possible implementation, before the computer device performs a first recognition process on the retrieval data based on the business category using the intent recognition model, and determines the intent category corresponding to the retrieval data from multiple intent categories corresponding to the business category, the method further includes: acquiring multiple sample data, which includes multiple sample retrieval data corresponding to different business categories and sample intent categories corresponding to each of the multiple sample retrieval data; training the initial intent recognition model based on the multiple sample data to obtain an intent recognition model including a target recognition layer, which is a recognition layer corresponding to multiple business categories; copying the target recognition layer according to the number of business categories to obtain a recognition layer corresponding to each business category; and training the recognition layer corresponding to each business category based on multiple sample retrieval data corresponding to that category and the sample intent categories corresponding to the multiple sample retrieval data. The model structure of this intent recognition model can be as follows: Figure 5 As shown.

[0112] Furthermore, in the embodiments of this application, when training the intent recognition model, the number of positive samples can be increased and the number of negative samples can be decreased to avoid low discrimination of the intent recognition model. This can be achieved by using one or more methods such as EDA, Mixup, or undersampling.

[0113] Optionally, when training the intent recognition model, the embodiments of this application can further enrich the information in the sample data. For example, the input to the intent recognition model can be sample retrieval data, or it can be sample retrieval data + sample intent category + high-frequency template of the sample retrieval data + matching information. The matching information indicates whether the sample intent category matches the sample retrieval data; that is, it indicates whether the sample data is a positive or negative sample. Compared to inputting only sample retrieval data, this can enrich the information to a certain extent and mitigate the impact of noise.

[0114] Optionally, this application embodiment employs a recall + fine-grained ranking approach when acquiring sample data. First, a coarse-grained segmentation is performed using recall, followed by a fine-grained segmentation using fine-grained ranking. This way, when subsequent business needs to expand to new intent categories, the recalled data can be further refined and categorized.

[0115] It should be noted that the embodiment of the present application can also determine the business category corresponding to the search data, and then determine the search object corresponding to the search data. Alternatively, the computer device performs third identification processing on the search data to obtain the search object of the search data, including: the computer device determines the target category based on the business category corresponding to the search data; the computer device determines the category of each character in the search data through an object recognition model; and the characters corresponding to the target category are determined as the search object corresponding to the search data.

[0116] For example, the business category corresponding to the search data is "tourism", the target category is "address", and the characters or words corresponding to the "address" component in the search data are determined as the search object.

[0117] 303、The computer device generates content items matching the search data based on the search intent information.

[0118] As can be seen from the above step 302, the search intent information includes at least one of the business category, the search intent category, and the search object. Therefore, when the computer device generates the content items matching the search data based on the search intent information, it can generate the content items matching the search data based on at least one of the business category, the intent category, and the search object.

[0119] The embodiments of the present application respectively provide a way to generate content items based on the business category, the intent category, and the search object. In one possible implementation, the computer device generates content items matching the search data based on the search intent information, including at least one of the following:

[0120] (1) Based on the intent category, the text data corresponding to the intent category is obtained from the first corresponding relationship, the first corresponding relationship is the corresponding relationship between the intent category and the text data, and the text data is the text data describing the content items; and based on the text data, the content items matching the search data are generated.

[0121] Alternatively, the text data is a text idea generated based on a template, the template is a template mined offline and matching the intent category. Alternatively, the text data is a text data mined offline and matching the intent category. Alternatively, the text data is made by a technician. The embodiments of the present application do not limit the source of the text data.

[0122] Alternatively, the text data includes title information and description information, and both the title information and the description information match the intent category. The first corresponding relationship is shown in Table 1.

[0123] Table 1

[0124]

[0125] It should be noted that the position of the retrieval object is reserved in the text data. For example, the "SCENIC scenic spot" in Table 1 is replaced by the retrieval object of the retrieval data.

[0126] (2) Based on the intention category, image data matching the intention category is obtained; and based on the image data, a content item matching the retrieval data is generated.

[0127] It should be noted that the content item in the embodiments of the present application can only include text data, or only include image data, or include both text data and image data. The embodiments of the present application do not limit the content item.

[0128] (3) In the case where the business category corresponding to the retrieval data is a target business category, based on the retrieval intention information, a content item matching the retrieval data is generated, and the target business category is a business category to which the content item to be recommended belongs.

[0129] For example, the business category corresponding to the retrieval data is "food", and the content item related to "food" is recommended to the user.

[0130] (4) Based on the retrieval object, a link of the retrieval object is obtained; and based on the link of the retrieval object, a content item matching the retrieval data is generated.

[0131] For example, the retrieval data input by the user is "how much is the ticket of scenic spot A", and the retrieval object of the retrieval data is "scenic spot A". Then, based on the ticket purchase link of "scenic spot A", a content item is generated and displayed to the user. The user can directly jump to the ticket purchase interface of "scenic spot A" through the content item, so that the user not only knows the ticket price of "scenic spot A", but also can directly purchase the ticket of "scenic spot A".

[0132] In some embodiments, the entity link is an entity link provided by an application. The application can provide multiple entity links. In order to find the retrieval corresponding entity link more conveniently and quickly, the application also provides a knowledge graph. The knowledge graph includes multiple entities, and the multiple entities are arranged according to progressive relationships. For example, the next level entities of the entity "xx city" in the knowledge graph include "xx city xx scenic spot", "xx city xx shopping center", "xx city xx square", and the like.

[0133] Each entity in the knowledge graph corresponds to an entity link. In the embodiments of the present application, the computer device finds an entity matching the retrieval object from the knowledge graph based on the retrieval object, and obtains the entity link corresponding to the entity. When the computer device finds an entity matching the retrieval object from the knowledge graph based on the retrieval object, the computer device can score multiple entities in the knowledge graph, and take the entity with the highest score as the entity matching the retrieval object.

[0134] Optionally, the following formula is used to obtain the score of the entity in the knowledge graph:

[0135] Wherein, f is the score of the entity, argmax is a function for finding the entity with the maximum score. s represents the word sequence of the search object, e represents the entity in the knowledge graph; q represents the search data, q-s represents the context of the search data, for example, the input time of the search data, etc. P represents the probability. E represents the knowledge graph, S q represents the set of word sequences of the search data. P(e|s) represents the entity score of the search object, that is, the probability that the search object refers to the entity. P(q-s|e) represents the context score of the entity.

[0136] For example, the search object is "B Park MMM", and the word sequence of the search object can be "B Park", "MMM", and "B Park MMM". The entities and entity scores related to these word sequences in the knowledge graph are shown in Table 2.

[0137] Table 2

[0138]

[0139]

[0140] Wherein, the click score is obtained based on the statistical user behavior.

[0141] "MMM Shopping Park" in Table 1 has the highest score, so "MMM Shopping Park" is taken as the link of the search object.

[0142] 304、The computer device displays the content item as the search result corresponding to the search data.

[0143] When displaying the content item, the computer device can display the content item at the first position, or at other positions, and the display mode of the content item is not limited by the embodiments of the application.

[0144] The content item processing method provided by the embodiments of the application generates a suitable content item by obtaining the search intent information of the search data, and delivers the content item based on the search data, which not only ensures the matching between the search data and the content item, but also avoids the situation that the content item is not delivered in the matched search data, thereby improving the delivery effect.

[0145] Figure 6 is a flowchart of a content item processing method provided by an embodiment of the application. The embodiments of the application take a computer device as an example of an execution subject, and the embodiments include:

[0146] 601、The computer device acquires a plurality of pieces of search data.

[0147] In an embodiment of the present application, the plurality of pieces of search data can be data input by a user for searching, or can be other data capable of representing information of interest to the user, that is, other data input by the user can be acquired as search data in the embodiment of the present application.

[0148] In a possible implementation manner, as shown in Figure 7 The computer device acquires a plurality of pieces of search data, including at least one of the following:

[0149] (1) Acquiring search data input by a user.

[0150] The search data can be search data input by the user in any application platform. For example, search data input by the user in application A; or for example, search data input by the user in application B, and the like. The embodiment of the present application does not limit the search data input by the user. The computer device can acquire the search data input by the user according to a search record, or can acquire the search data input by the user through other manners.

[0151] (2) Acquiring an existing bid word, and taking the existing bid word as search data.

[0152] The existing bid word can be an existing bid word in the Internet, or can be a bid word being used by a target application, or can be a bid word being used by a competing application, and the embodiment of the present application does not limit the existing bid word.

[0153] (3) Acquiring note data input by a user, and acquiring search data based on the note data.

[0154] The user can post a note in the target application, and the content of the note can be food recommendation, scenic spot recommendation, scenic spot strategy, and the like, and the embodiment of the present application does not limit the content of the note. It is considered that the content recorded in the note is not only content that the user wants to share, but also content that other users want to understand, and therefore, search data can be acquired based on the note data. For example, a note records a travel strategy of city A, and therefore, search data “what is interesting in city A” can be acquired based on the note.

[0155] (4) Acquiring list data, and determining the list data as search data.

[0156] A list module is provided in the target application, and the list module is used to provide a plurality of pieces of most popular information in the target application, and the plurality of pieces of information are arranged according to the degree of popularity. Therefore, the list data in the list module is data that the user is very interested in, and therefore, the list data can be determined as search data.

[0157] (5) Obtain topic data, and determine the topic data as the retrieval data.

[0158] The topic module is arranged in the target application, and the topic module is configured to provide a plurality of hottest topics in the target application, and the plurality of topics are arranged according to the degree of popularity. Therefore, the topic data in the topic module is data that is very concerned by the user, and therefore, the topic data can be determined as the retrieval data.

[0159] (6) Obtain a merchant graph, and obtain the retrieval data based on the merchant graph.

[0160] The merchant graph is configured to record a plurality of merchants that are registered in the target application, and each merchant provides a corresponding service. The user can search for a related merchant, and purchase a related service provided by the merchant in the target application. Therefore, the plurality of merchants registered in the target application can be a search target of the user, and therefore, the retrieval data can be obtained based on the merchant graph.

[0161] (7) Obtain user original content data, and obtain the retrieval data based on the user original content data.

[0162] The user can publish original content data in the target application, and the original content data can include content that is interested by other users. Since the retrieval data is configured to retrieve content that is interested by the user, the retrieval data can be obtained based on the user original content data.

[0163] In a possible implementation manner, the computer device obtains a plurality of retrieval data, including: obtaining the plurality of retrieval data from the Internet. Optionally, the content item processing method provided by the embodiments of the present application is applicable to processing a content item provided by a target application. If the obtained retrieval data is irrelevant to the business of the target application, the retrieval data can be discarded. As shown in Figure 8 After the retrieval data is obtained, it is determined whether the retrieval intention of the retrieval data is strongly related, weakly related or unrelated to the target application.

[0164] For example, the first application is an e-commerce application, and a plurality of merchants are registered in the first application; and the second application is a retrieval platform. The first application can obtain a plurality of retrieval data from the Internet. If the retrieval intention of the retrieval data is related to the business of the first application, the retrieval data is retained; and if the retrieval intention of the retrieval data is irrelevant to the business of the first application (for example, the retrieval data is configured to retrieve a song), the retrieval data is discarded.

[0165] 602. The computer device performs identification processing on the plurality of retrieval data, to obtain retrieval intention information of each retrieval data.

[0166] The step 602 is the same as the step 302 described above, and will not be described again.

[0167] 603、The computer device determines at least one piece of search data corresponding to different search intention information based on the search intention information of each piece of search data.

[0168] People can use different expressions when expressing a certain meaning, which results in different search data corresponding to the same search intention information. The above step 603 actually classifies the search data based on the search intention information of each piece of search data, so that the search data corresponding to the same search intention information is classified together.

[0169] 604、The computer device generates a content item matching the search intention information based on each search intention information in the different search intention information.

[0170] The above step 604 is the same as the above step 304, and will not be repeated here.

[0171] 605、The computer device determines at least one piece of search data corresponding to the search intention information as the delivery data of the content item matching the search intention information.

[0172] That is, in the embodiment of the application, after obtaining the content item matching a certain search intention information, all search data corresponding to the search intention information are determined as the delivery data of the content item. In this way, when the search data input by the user matches any delivery data corresponding to the search intention information, it can be determined that the search intention of the user is the search intention represented by the search intention information, and the content item matching the search intention information can be recommended to the user.

[0173] 606、The computer device acquires the input search data in response to a search operation.

[0174] The search operation is an operation for triggering the computer device to search, which can be any one or a combination of multiple operations such as a click operation, a sliding operation, a single-click operation, a double-click operation, etc. The embodiment of the application does not limit the search operation.

[0175] 607、The computer device displays the content item corresponding to the delivery data when displaying the search result of the search data in the case that the input search data matches the delivery data.

[0176] In the embodiment of the application, after the search data is acquired, the search data is matched with multiple delivery data, the delivery data matching the search data is determined, and the content item corresponding to the delivery data is displayed to the user.

[0177] The content item processing method provided in the embodiments of the present application classifies a large amount of search data according to search intention information of the search data, and uses search data corresponding to the same search intention information as delivery data of a content item matched with the search intention information. In this way, the corresponding content item can be accurately obtained as long as the search data input by the user matches the delivery data. Since the delivery data is obtained from a large amount of search data on the Internet, the delivery data is rich and diverse, so that the delivery data matched with the input search data can be more accurately obtained, thereby providing the user with more accurate content items.

[0178] Figure 9 is a structural schematic diagram of a content item processing apparatus provided in the embodiments of the present application, referring to Figure 9 The apparatus comprises:

[0179] An identification module 901 is configured to perform identification processing on search data to obtain search intention information of the search data, wherein the search intention information is used to indicate a search intention of the search data.

[0180] A generation module 902 is configured to generate a content item matched with the search data based on the search intention information.

[0181] A delivery module 903 is configured to deliver the content item based on the search data.

[0182] As shown in Figure 10 In a possible implementation manner, the identification module 901 is configured to perform at least one of the following:

[0183] The search intention information comprises an intention category corresponding to the search data, wherein the intention category is used to indicate a category to which the search intention of the search data belongs; the search data is subjected to first identification processing to obtain the intention category corresponding to the search data.

[0184] The search intention information comprises a business category corresponding to the search data, wherein the business category indicates a business category to which a search object of the search data belongs; the search data is subjected to second identification processing to obtain the business category corresponding to the search data.

[0185] The search intention information comprises a search object of the search data; the search data is subjected to third identification processing to obtain the search object of the search data.

[0186] In a possible implementation manner, the identification module 901 is configured to perform first identification processing on the search data based on a business category corresponding to the search data, and determine the intention category corresponding to the search data from a plurality of intention categories corresponding to the business category.

[0187] In a possible implementation, the identification module 901 includes:

[0188] The acquisition unit 9011 is configured to acquire an intent identification model, the intent identification model including a feature extraction layer, a feature processing layer, and an identification layer corresponding to each business category, the identification layer corresponding to each business category being configured to determine, from a plurality of intent categories corresponding to the business category, an intent category to which a retrieval intent of retrieval data belongs;

[0189] The first processing unit 9012 is configured to perform feature extraction on the retrieval data through the feature extraction layer to obtain first feature data;

[0190] The second processing unit 9013 is configured to perform processing on the first feature data through the feature extraction layer to obtain second feature data;

[0191] The input unit 9014 is configured to input the second feature data to the identification layer corresponding to the business category corresponding to the retrieval data based on the business category corresponding to the retrieval data;

[0192] The third processing unit 9015 is configured to perform identification processing on the second feature data through the identification layer to obtain an intent category corresponding to the retrieval data.

[0193] In a possible implementation, the first processing unit 9012 is configured to perform word segmentation processing on the retrieval data to obtain a plurality of word characters;

[0194] The first processing unit 9012 is configured to perform feature extraction on each word character in the plurality of word characters through the feature extraction layer to obtain first feature data corresponding to each word character;

[0195] The second processing unit 9013 is configured to, for the first feature data corresponding to each word character, perform fusion processing on the first feature data corresponding to the word character and first feature data of other word characters in the plurality of word characters except the word character through the feature extraction layer to obtain second feature data of the word character.

[0196] In a possible implementation, the apparatus further includes a training module 904.

[0197] The training module 904 is configured to acquire a plurality of sample data, the plurality of sample data including a plurality of sample retrieval data corresponding to different business categories and sample intent categories corresponding to the plurality of sample retrieval data respectively;

[0198] The training module 904 is configured to train the initial intent recognition model based on the plurality of sample data to obtain an intent recognition model comprising a target recognition layer, the target recognition layer being a recognition layer corresponding to a plurality of service categories;

[0199] The training module 904 is configured to replicate the target recognition layer according to the number of service categories to obtain a recognition layer corresponding to each service category.

[0200] The training module 904 is configured to train the recognition layer corresponding to each service category based on a plurality of sample search data corresponding to the category and a sample intent category corresponding to the plurality of sample search data.

[0201] In a possible implementation, the generation module 902 is configured to perform at least one of the following:

[0202] Based on the intent category, obtain text data corresponding to the intent category from a first correspondence relationship, the first correspondence relationship being a correspondence relationship between an intent category and text data, the text data being text data describing a content item; and based on the text data, generate a content item matching the search data;

[0203] Based on the intent category, obtain image data matching the intent category; and based on the image data, generate a content item matching the search data;

[0204] In a case where the service category corresponding to the search data is a target service category, based on the search intent information, generate a content item matching the search data, the target service category being a service category to which the recommended content item belongs;

[0205] Based on the search object, obtain a link of the search object; and based on the link of the search object, generate a content item matching the search data.

[0206] In a possible implementation, the search data is search data obtained in response to a search data acquisition, and the delivery module 903 is configured to display the content item as a search result corresponding to the search data.

[0207] In a possible implementation, the search data is search data obtained from a search record, and the delivery module 903 comprises:

[0208] A determination unit 9031 is configured to determine the search data as delivery data of the content item.

[0209] An acquisition unit 9032 is configured to acquire input search data in response to a search operation.

[0210] The display unit 9033 is used to display the content item when displaying the search results of the search data, provided that the input search data matches the delivery data.

[0211] In one possible implementation, the identification module 901 is used to identify and process multiple search data to obtain search intent information for each search data; based on the search intent information for each search data, at least one search data corresponding to different search intent information is determined;

[0212] The generation module 902 is used to determine at least one piece of search data corresponding to the search intent information as the delivery data of the content item that matches the search intent information.

[0213] It should be noted that the content item processing apparatus provided in the above embodiments is only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the computer device can be divided into different functional modules to complete all or part of the functions described above. In addition, the content item processing apparatus and the content item processing method embodiments provided in the above embodiments belong to the same concept, and their specific implementation process can be found in the method embodiments, which will not be repeated here.

[0214] In an exemplary embodiment, a computer device is provided, the computer device including one or more processors and one or more memories, the one or more memories storing at least one piece of program code, the at least one piece of program code being loaded and executed by the one or more processors to implement the content item processing method as described in the above embodiments.

[0215] Optionally, the computer device is provided as a terminal. Figure 11 A structural block diagram of a terminal 1100 provided in an exemplary embodiment of this application is shown. The terminal 1100 may be a smartphone, tablet computer, MP3 player (Moving Picture Experts Group Audio Layer III), MP4 player (Moving Picture Experts Group Audio Layer IV), laptop computer, or desktop computer. The terminal 1100 may also be referred to as a user device, portable terminal, laptop terminal, desktop terminal, or other names.

[0216] Terminal 1100 includes a processor 1101 and a memory 1102.

[0217] The processor 1101 can include one or more processing cores, such as a 4-core processor, an 8-core processor, and the like. The processor 1101 can be implemented in at least one of a hardware form of a DSP (Digital Signal Processing), an FPGA (Field-Programmable Gate Array), a PLA (Programmable Logic Array). The processor 1101 can also include a main processor and a coprocessor, the main processor being a processor for processing data in an awake state, also known as a CPU (Central Processing Unit), and the coprocessor being a low-power processor for processing data in a standby state. In some embodiments, the processor 1101 can be integrated with a GPU (Graphics Processing Unit) that is responsible for rendering and drawing the content required to be displayed by the display screen. In some embodiments, the processor 1101 can also include an AI (Artificial Intelligence) processor for processing machine learning-related computing operations.

[0218] The memory 1102 can include one or more computer-readable storage media that can be non-transitory. The memory 1102 can also include a high-speed random access memory, and a nonvolatile memory such as one or more disk storage devices, flash storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 1102 is used to store at least one program code for being executed by the processor 1101 to implement the content processing method provided by the method embodiments in the present application.

[0219] In some embodiments, the terminal 1100 can also optionally include a peripheral device interface 1103 and at least one peripheral device. The processor 1101, the memory 1102, and the peripheral device interface 1103 can be connected through a bus or a signal line. Each peripheral device can be connected to the peripheral device interface 1103 through a bus, a signal line, or a circuit board. Specifically, the peripheral device includes at least one of a radio frequency circuit 1104, a display screen 1105, a camera 1106, an audio circuit 1107, a positioning component 1108, and a power supply 1109.

[0220] The peripheral interface 1103 can be used to connect at least one peripheral device related to I / O (Input / Output) to the processor 1101 and the memory 1102. In some embodiments, the processor 1101, the memory 1102 and the peripheral interface 1103 are integrated on the same chip or circuit board; in some other embodiments, any one or two of the processor 1101, the memory 1102 and the peripheral interface 1103 can be implemented on a separate chip or circuit board, and the present embodiment is not limited in this regard.

[0221] The radio frequency circuit 1104 is used to receive and transmit RF (Radio Frequency) signals, also known as electromagnetic signals. The radio frequency circuit 1104 communicates with a communication network and other communication devices through electromagnetic signals. The radio frequency circuit 1104 converts electrical signals into electromagnetic signals for transmission, or converts received electromagnetic signals into electrical signals. Optionally, the radio frequency circuit 1104 includes an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a subscriber identity module card, and the like. The radio frequency circuit 1104 can communicate with other terminals through at least one wireless communication protocol. The wireless communication protocol includes but is not limited to: a metropolitan area network, various generations of mobile communication networks (2G, 3G, 4G and 5G), a wireless local area network and / or a WiFi (Wireless Fidelity) network. In some embodiments, the radio frequency circuit 1104 can also include NFC (Near Field Communication) related circuitry, and the present application is not limited in this regard.

[0222] The display screen 1105 is configured to display a UI (User Interface). The UI can include graphics, text, icons, video, and any combination thereof. When the display screen 1105 is a touch display screen, the display screen 1105 is further configured to capture touch signals on or above the surface of the display screen 1105. The touch signals can be input to the processor 1101 as control signals for processing. In this case, the display screen 1105 can also be configured to provide virtual buttons and / or virtual keyboard, also known as soft buttons and / or soft keyboard. In some embodiments, the display screen 1105 can be one, arranged on the front panel of the terminal 1100; in other embodiments, the display screen 1105 can be at least two, arranged on different surfaces of the terminal 1100 or in a folding design; in still other embodiments, the display screen 1105 can be a flexible display screen, arranged on a curved surface or a folding surface of the terminal 1100. Even, the display screen 1105 can also be arranged in an irregular shape other than a rectangle, i.e., a special-shaped screen. The display screen 1105 can be made of materials such as LCD (Liquid Crystal Display) or OLED (Organic Light-Emitting Diode).

[0223] The camera assembly 1106 is configured to capture images or videos. Optionally, the camera assembly 1106 includes a front-facing camera and a rear-facing camera. The front-facing camera is arranged on the front panel of the terminal, and the rear-facing camera is arranged on the back of the terminal. In some embodiments, the rear-facing camera is at least two, which is any one of a main camera, a depth-of-field camera, a wide-angle camera, and a telephoto camera, to realize the background blur function of the main camera and the depth-of-field camera, the panoramic shooting and VR (Virtual Reality) shooting function of the main camera and the wide-angle camera, or other fusion shooting functions. In some embodiments, the camera assembly 1106 can further include a flash. The flash can be a single-color temperature flash or a dual-color temperature flash. The dual-color temperature flash refers to the combination of a warm light flash and a cold light flash, which can be used for light compensation under different color temperatures.

[0224] The audio circuit 1107 can include a microphone and a speaker. The microphone is used to collect sound waves of a user and an environment, and convert the sound waves into an electrical signal input to the processor 1101 for processing, or input to the radio frequency circuit 1104 to realize voice communication. For the purpose of stereo sound collection or noise reduction, the microphone can be multiple, respectively arranged at different parts of the terminal 1100. The microphone can also be an array microphone or an omnidirectional collection type microphone. The speaker is used to convert an electrical signal from the processor 1101 or the radio frequency circuit 1104 into sound waves. The speaker can be a traditional diaphragm speaker, or a piezoelectric ceramic speaker. When the speaker is a piezoelectric ceramic speaker, not only can the electrical signal be converted into a sound wave audible to humans, but also can be converted into a sound wave inaudible to humans for ranging purposes. In some embodiments, the audio circuit 1107 can also include a headphone jack.

[0225] The positioning component 1108 is used to position the current geographic position of the terminal 1100 to realize navigation or LBS (Location Based Service). The positioning component 1108 can be a positioning component based on the GPS (Global Positioning System) of the United States, the Beidou system of China, the Glonass system of Russia or the Galileo system of the European Union.

[0226] The power supply 1109 is used to supply power to each component in the terminal 1100. The power supply 1109 can be alternating current, direct current, disposable battery or rechargeable battery. When the power supply 1109 includes a rechargeable battery, the rechargeable battery can support wired charging or wireless charging. The rechargeable battery can also be used to support fast charging technology.

[0227] In some embodiments, the terminal 1100 also includes one or more sensors 1110. The one or more sensors 1110 include but are not limited to: an acceleration sensor 1111, a gyroscope sensor 1112, a pressure sensor 1113, a fingerprint sensor 1114, an optical sensor 1115 and a proximity sensor 1116.

[0228] The acceleration sensor 1111 can detect the acceleration magnitude in three coordinate axes of the coordinate system established by the terminal 1100. For example, the acceleration sensor 1111 can be used to detect the components of the gravitational acceleration in three coordinate axes. The processor 1101 can control the display screen 1105 to display the user interface in a landscape view or a portrait view according to the gravitational acceleration signal collected by the acceleration sensor 1111. The acceleration sensor 1111 can also be used for game or user motion data collection.

[0229] The gyroscope sensor 1112 can detect the body direction and rotation angle of the terminal 1100, and can collect 3D motions of the user on the terminal 1100 in cooperation with the acceleration sensor 1111. The processor 1101 can implement the following functions according to the data collected by the gyroscope sensor 1112: motion sensing (e.g., changing a UI according to a tilt operation of the user), image stabilization during shooting, game control, and inertial navigation.

[0230] The pressure sensor 1113 can be disposed on the side frame of the terminal 1100 and / or the lower layer of the display screen 1105. When the pressure sensor 1113 is disposed on the side frame of the terminal 1100, the grip signal of the user on the terminal 1100 can be detected, and the left-hand / right-hand recognition or shortcut operation can be performed by the processor 1101 according to the grip signal collected by the pressure sensor 1113. When the pressure sensor 1113 is disposed on the lower layer of the display screen 1105, the operable control on the UI interface can be controlled by the processor 1101 according to the pressure operation of the user on the display screen 1105. The operable control includes at least one of a button control, a scroll bar control, an icon control, and a menu control.

[0231] The fingerprint sensor 1114 is used to collect the fingerprint of the user, and the identity of the user can be recognized by the processor 1101 according to the fingerprint collected by the fingerprint sensor 1114, or by the fingerprint sensor 1114 according to the collected fingerprint. When the identity of the user is recognized as a trusted identity, the processor 1101 authorizes the user to perform a related sensitive operation, which includes unlocking the screen, viewing encrypted information, downloading software, payment, and changing settings, etc. The fingerprint sensor 1114 can be disposed on the front, back, or side of the terminal 1100. When the physical button or the manufacturer's logo is disposed on the terminal 1100, the fingerprint sensor 1114 can be integrated with the physical button or the manufacturer's logo.

[0232] The optical sensor 1115 is used to collect the ambient light intensity. In one embodiment, the processor 1101 can control the display brightness of the display screen 1105 according to the ambient light intensity collected by the optical sensor 1115. Specifically, when the ambient light intensity is high, the display brightness of the display screen 1105 is increased, and when the ambient light intensity is low, the display brightness of the display screen 1105 is decreased. In another embodiment, the processor 1101 can also dynamically adjust the shooting parameters of the camera assembly 1106 according to the ambient light intensity collected by the optical sensor 1115.

[0233] The proximity sensor 1116, also referred to as a distance sensor, is arranged on the front panel of the terminal 1100. The proximity sensor 1116 is configured to collect the distance between the user and the front of the terminal 1100. In an embodiment, when the proximity sensor 1116 detects that the distance between the user and the front of the terminal 1100 gradually decreases, the display screen 1105 is switched from the bright screen state to the screen-off state under the control of the processor 1101; when the proximity sensor 1116 detects that the distance between the user and the front of the terminal 1100 gradually increases, the display screen 1105 is switched from the screen-off state to the bright screen state under the control of the processor 1101.

[0234] Those skilled in the art can understand that the structure shown in the foregoing embodiments is not a limitation on the terminal 1100, and the terminal 1100 can include more or fewer components than those shown in the drawings, or combine certain components, or adopt a different arrangement of components. Figure 11

[0235] Optionally, the computer device is provided as a server. Figure 12 FIG. 12 is a structural diagram of a server according to an embodiment of the present application. The server 1200 can have a large difference due to different configurations or performances, and can include one or more processors (Central Processing Units, CPUs) 1201 and one or more memories 1202. The memory 1202 stores at least one program code, which is loaded and executed by the processor 1201 to implement the method provided by the above-mentioned various method embodiments. Of course, the server can also have a wired or wireless network interface, a keyboard, an input and output interface, and other components for implementing device functions, which are not described herein.

[0236] The server 1200 is configured to execute the steps performed by the server in the above-mentioned method embodiments.

[0237] In an exemplary embodiment, a computer readable storage medium, such as a memory including program code, is also provided, which can be executed by a processor in a computer device to complete the content item processing method in the above-mentioned embodiments. For example, the computer readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.

[0238] In an exemplary embodiment, a computer program or computer program product is also provided, which includes computer program code. When the computer program code is executed by a computer, the computer implements the content item processing method in the above-mentioned embodiments.

[0239] ​Those skilled in the art can understand that all or part of the steps of the above-mentioned embodiments can be completed by hardware, or can be instructed to relevant hardware by program. The program can be stored in a computer readable storage medium, and the storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0240] The above only describes optional embodiments of the present application and is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for processing content items, characterized in that, The method includes: Multiple search data entries are obtained from the search records; the multiple search data entries are identified and processed to obtain the search intent information for each search data entry, the search intent information including the business category corresponding to the search data entry, the intent category corresponding to the search data entry, and the search object of the search data entry; Based on the search intent information of each search data, the search data corresponding to each search intent information is determined, and the search data corresponding to each search intent information has the same search intent. Based on each search intent, generate content items that match the search intent; determine the search data corresponding to the search intent as the delivery data for the content items that match the search intent; In response to a search operation, the input search data is obtained; if the input search data matches the delivery data, the content items corresponding to the delivery data are displayed when displaying the search results of the search data. The step of identifying and processing any search data to obtain the search intent information of the search data includes: performing a second identification process on the search data to obtain the business category corresponding to the search data; determining a target component category based on the business category corresponding to the search data; determining the component category of each character in the search data through an object recognition model; determining the character corresponding to the target component category as the search object corresponding to the search data; and performing a first identification process on the search data based on the business category corresponding to the search data to determine the intent category corresponding to the search data from multiple intent categories corresponding to the business category. The step of generating a content item matching any search intent information includes at least one of the following: based on the intent category in the search intent information, obtaining text data corresponding to the intent category from a first correspondence, wherein the first correspondence is a correspondence between intent categories and text data, and the text data is text data describing the content item; generating a content item matching the search intent information based on the text data; obtaining image data matching the intent category based on the intent category; generating a content item matching the search intent information based on the image data; if the business category in the search intent information is a target business category, generating a content item matching the search intent information based on the search intent information, wherein the target business category is the business category to which the content item to be recommended belongs; obtaining the link of the search object based on the search intent information; and generating a content item matching the search intent information based on the link of the search object.

2. The method according to claim 1, characterized in that, The first identification process, based on the business category corresponding to the search data and using an intent recognition model, involves determining the intent category corresponding to the search data from multiple intent categories corresponding to the business category. This includes: The intent recognition model is obtained, which includes a feature extraction layer, a feature processing layer, and an identification layer corresponding to each business category. The identification layer corresponding to each business category is used to determine the intent category to which the retrieval intent of the retrieval data belongs from multiple intent categories corresponding to the business category. The feature extraction layer extracts features from the retrieved data to obtain first feature data. The first feature data is processed through the feature extraction layer to obtain the second feature data; Based on the business category corresponding to the retrieved data, the second feature data is input into the recognition layer corresponding to the business category; The recognition layer processes the second feature data to obtain the intent category corresponding to the retrieved data.

3. The method according to claim 2, characterized in that, The step of extracting features from the retrieved data through the feature extraction layer to obtain first feature data includes: The retrieved data is segmented into words to obtain multiple word characters; Through the feature extraction layer, features are extracted from each of the multiple words to obtain the first feature data corresponding to each word; The step of processing the first feature data through the feature extraction layer to obtain the second feature data includes: For each word, the first feature data is fused with the first feature data of other words in the plurality of words to obtain the second feature data of the word.

4. The method according to claim 2, characterized in that, Before determining the intent category corresponding to the search data from multiple intent categories corresponding to the business category by performing a first identification process on the search data based on the business category and through the intent recognition model, the method further includes: Acquire multiple sample data, which include multiple sample retrieval data corresponding to different business categories and sample intent categories corresponding to the multiple sample retrieval data respectively; Based on the multiple sample data, the initial intent recognition model is trained to obtain an intent recognition model including a target recognition layer, wherein the target recognition layer is a recognition layer corresponding to multiple service categories; Based on the number of business categories, the target recognition layer is copied to obtain the recognition layer corresponding to each business category; For each business category, the recognition layer is trained based on multiple sample retrieval data corresponding to the category and the sample intent category corresponding to the multiple sample retrieval data.

5. A content item processing apparatus, characterized in that, The device includes: The identification module is used to obtain multiple search data from the search records; to perform identification processing on the multiple search data to obtain the search intent information of each search data, the search intent information including the business category corresponding to the search data, the intent category corresponding to the search data, and the search object of the search data; and to summarize based on the search intent information of each search data to determine the search data corresponding to each search intent information, and the search data corresponding to each search intent information has the same search intent. The generation module is used to generate content items that match each search intent information based on the search intent information; and to determine the search data corresponding to the search intent information as the delivery data of the content items that match the search intent information. The delivery module is used to respond to a search operation, obtain the input search data, and when the input search data matches the delivery data, display the content items corresponding to the delivery data while displaying the search results of the search data. The identification module is configured to perform a second identification process on the search data to obtain the business category corresponding to the search data; determine a target component category based on the business category corresponding to the search data; determine the component category of each character in the search data through an object recognition model; identify the character corresponding to the target component category as the search object corresponding to the search data; and perform a first identification process on the search data based on the business category corresponding to the search data to determine the intent category corresponding to the search data from multiple intent categories corresponding to the business category. The generation module is configured to: obtain text data corresponding to the intent category from a first correspondence relationship based on the intent category in the retrieval intent information, wherein the first correspondence relationship is a correspondence between intent categories and text data, and the text data is text data describing content items; generate content items matching the retrieval intent information based on the text data; obtain image data matching the intent category based on the intent category; generate content items matching the retrieval intent information based on the image data; if the business category in the retrieval intent information is a target business category, generate content items matching the retrieval intent information based on the retrieval intent information, wherein the target business category is the business category to which the content item to be recommended belongs; obtain the link of the retrieval object based on the retrieval object in the retrieval intent information; and generate content items matching the retrieval intent information based on the link of the retrieval object.

6. A computer device, characterized in that, The computer device includes one or more processors and one or more memories, wherein at least one piece of program code is stored in the one or more memories, and the at least one piece of program code is loaded and executed by the one or more processors to perform the operations performed by the content item processing method as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The storage medium stores at least one piece of program code, which is loaded and executed by a processor to perform the operations performed by the content item processing method as described in any one of claims 1 to 4.

Images