Data processing method, device, electronic device and storage medium

Abstract

The present invention provides a data processing method, device, electronic device and storage medium. For each type of manipulation behavior supported by the current object manipulated by the user, the current manipulation behavior data of the user on the current object is obtained. For each type of manipulation behavior supported by the historical objects manipulated by the user in the historical process, the historical manipulation behavior data of the user on the historical objects is obtained. Based on the current manipulation behavior data and the historical manipulation behavior data, the user's emotion classification of the current object is obtained. Through this application, the user's emotion classification of the current object is obtained based on the historical manipulation behavior data of the user on the historical objects in the historical process, combined with the user's current manipulation behavior data of the current object, so that the user's emotion classification of the current object can be obtained according to the user's actual situation (such as behavioral habits, etc.), and the accuracy of the obtained user's emotion classification of the current object can be improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a data processing method, device, electronic equipment and storage medium. Background Art

[0002] In personalized recommendation systems, users’ sentiment classification of objects is often obtained through their feedback on the objects. The user’s sentiment classification of the objects and the attributes of the objects are then combined to update the user’s preference information. Then, objects that meet the user’s preferences are recommended to the user based on the updated user’s preference information.

[0003] Among them, user feedback is often the user's manipulation of the object, such as likes, dislikes, purchases, and shares.

[0004] In this way, the user's emotional classification of the object can be determined based on the user's manipulation behavior on the object. However, the inventors have found that this is likely to result in inaccurate emotional classification of the user on the object. Summary of the invention

[0005] The present application illustrates a data processing method, device, electronic device and storage medium.

[0006] In a first aspect, the present application shows a data processing method, which is applied to a terminal, and the method includes:

[0007] When a user manipulates a current object, for each manipulation type supported by the current object, obtaining current manipulation behavior data of the user on the current object, wherein the current manipulation behavior data includes the manipulation type of the manipulation behavior actually performed by the user on the current object and the manipulation type of the manipulation behavior not performed;

[0008] For each historical object manipulated by the user in the historical process, for each manipulation type supported by the historical object, obtaining the historical manipulation behavior data of the user on the historical object, the historical manipulation behavior data including the manipulation type of the manipulation behavior actually performed by the user on the historical object and the manipulation type of the manipulation behavior not performed;

[0009] The emotion classification of the user towards the current object is obtained according to the current manipulation behavior data and the historical manipulation behavior data.

[0010] In an optional implementation, obtaining the user's emotion classification of the current object according to the current manipulation behavior data and the historical manipulation behavior data includes:

[0011] Acquire a sentiment classification acquisition model, wherein the sentiment classification acquisition model is obtained by training based on a plurality of training data sets, wherein the training data sets include: first sample manipulation behavior data of a sample user on a first sample object, second sample manipulation behavior data of the sample user on a second sample object, and annotated sentiment classification of the sample user on the second sample object;

[0012] Inputting the current manipulation behavior data and the historical manipulation behavior data into the emotion classification acquisition model, so that the emotion classification acquisition model processes the current manipulation behavior data and the historical manipulation behavior data, obtains the user's emotion classification of the current object, and outputs the user's emotion classification of the current object;

[0013] Acquire the sentiment classification of the user to the current object output by the sentiment classification acquisition model.

[0014] In an optional implementation, acquiring the sentiment classification acquisition model includes:

[0015] According to the historical manipulation behavior data, an emotion classification acquisition model suitable for the user is acquired from a plurality of trained emotion classification acquisition models.

[0016] In an optional implementation, acquiring a sentiment classification acquisition model suitable for the user from a plurality of trained sentiment classification acquisition models according to the historical manipulation behavior data includes:

[0017] Determining, from a plurality of user sets, a user set to which the user belongs according to the historical manipulation behavior data;

[0018] Among the sentiment classification acquisition models respectively applicable to each user set, a sentiment classification acquisition model applicable to the user set to which the user belongs is selected.

[0019] In an optional implementation, obtaining the user's emotion classification of the current object according to the current manipulation behavior data and the historical manipulation behavior data includes:

[0020] For any manipulation behavior of a manipulation type supported by the current object, determine the current execution status of the manipulation behavior of the current object with respect to the manipulation type by the user according to the current manipulation behavior data, and determine the historical execution status of the manipulation behavior of the historical object with respect to the manipulation type by the user according to the historical manipulation behavior data, and obtain the emotion score of the user with respect to the manipulation behavior of the current object with respect to the manipulation type according to the current execution status and the historical execution status;

[0021] The emotion classification of the user on the current object is obtained according to the emotion scores of the user's manipulation behaviors on the current object regarding each manipulation type.

[0022] In an optional implementation, obtaining the emotion score of the user on the manipulation behavior of the current object with respect to the manipulation type according to the current execution status and the historical execution status includes:

[0023] Obtaining a basic emotion classification of the manipulation behavior of the manipulation type;

[0024] Counting statistical results of whether the user has an execution tendency to perform a manipulation behavior of the manipulation type on the historical object according to the historical execution situation, and determining an execution result of whether the user has performed a manipulation behavior of the manipulation type on the current object according to the current execution situation;

[0025] The emotion score of the user for the manipulation behavior of the current object with respect to the manipulation type is obtained according to the basic emotion classification of the manipulation behavior of the manipulation type, the statistical result and the execution result.

[0026] In an optional implementation, the statistical result of counting, according to the historical execution situation, whether the user has an execution tendency to execute the manipulation behavior of the manipulation type on the historical object includes:

[0027] Acquire the historical number of historical objects manipulated by the user in the historical process, and acquire the number of times the user performs manipulation actions of the manipulation type on the historical objects in the process of manipulating the historical objects in the historical process;

[0028] A statistical result of whether the user has an execution tendency to execute the manipulation behavior of the manipulation type on the historical object is counted according to the history quantity and the execution times.

[0029] In an optional implementation, the statistical result of counting, based on the history quantity and the execution times, whether the user has an execution tendency to execute the manipulation behavior of the manipulation type on the history object includes:

[0030] Calculating a ratio between the number of executions and the number of history;

[0031] When the ratio is greater than or equal to a preset ratio, determining that the user has an execution tendency to execute a manipulation behavior of the manipulation type on the historical object;

[0032] or,

[0033] When the ratio is smaller than a preset ratio, it is determined that the user does not have an execution tendency to execute a manipulation behavior of the manipulation type on the historical object.

[0034] In an optional implementation, obtaining the user's emotion score for the manipulation behavior of the current object with respect to the manipulation type according to the basic emotion classification of the manipulation behavior of the manipulation type, the statistical result, and the execution result includes:

[0035] In the correspondence between the basic emotion classification of the manipulation behavior, the statistical results of the execution tendency of the manipulation behavior, the execution results of the manipulation behavior and the user's emotion score for the manipulation behavior, find the user's emotion score for the manipulation behavior that corresponds to the basic emotion classification of the manipulation behavior of the manipulation type, the statistical results and the execution results, and use it as the user's emotion score for the manipulation behavior of the current object regarding the manipulation type.

[0036] In an optional implementation, obtaining the user's emotion classification for the current object according to the emotion scores of the user's manipulation behaviors for each manipulation type for the current object includes:

[0037] Taking a weighted sum of the emotion scores of the user's manipulation behaviors on the current object for each manipulation type, to obtain a total emotion score of the user on the current object;

[0038] The emotion classification of the user towards the current object is obtained according to the total emotion value of the user towards the current object.

[0039] In an optional implementation, obtaining the user's emotion classification for the current object according to the user's total emotion value for the current object includes:

[0040] In a case where the total emotion value of the user towards the current object is greater than a preset emotion threshold, determining that the emotion of the user towards the current object is classified as a positive emotion;

[0041] or,

[0042] When the total emotion value of the user towards the current object is equal to a preset emotion threshold, determining that the emotion of the user towards the current object is classified as a neutral emotion;

[0043] or,

[0044] When the total value of the user's emotion toward the current object is less than a preset emotion threshold, it is determined that the user's emotion toward the current object is classified as a negative emotion.

[0045] In a second aspect, the present application shows a data processing device, which is applied to a terminal, and the device includes:

[0046] A first acquisition module is used to acquire, when a user manipulates a current object, current manipulation behavior data of the user on the current object for each manipulation type supported by the current object, wherein the current manipulation behavior data includes the manipulation type of the manipulation behavior actually performed by the user on the current object and the manipulation type of the manipulation behavior not performed;

[0047] A second acquisition module is used to acquire, for each historical object that the user has manipulated in the historical process, historical manipulation behavior data of the user on the historical object for each manipulation type supported by the historical object, wherein the historical manipulation behavior data includes the manipulation type of the manipulation behavior actually performed by the user on the historical object and the manipulation type of the manipulation behavior not performed;

[0048] The third acquisition module is used to obtain the user's emotion classification of the current object according to the current manipulation behavior data and the historical manipulation behavior data.

[0049] In an optional implementation, the third acquisition module includes:

[0050] A first acquisition unit is used to acquire a sentiment classification acquisition model, wherein the sentiment classification acquisition model is obtained by training based on a plurality of training data sets, wherein the training data sets include: first sample manipulation behavior data of a sample user on a first sample object, second sample manipulation behavior data of the sample user on a second sample object, and a labeled sentiment classification of the sample user on the second sample object;

[0051] An input unit, used to input the current manipulation behavior data and the historical manipulation behavior data into the emotion classification acquisition model, so that the emotion classification acquisition model processes the current manipulation behavior data and the historical manipulation behavior data to obtain the user's emotion classification of the current object, and output the user's emotion classification of the current object;

[0052] The second acquisition unit is used to acquire the sentiment classification of the user to the current object output by the sentiment classification acquisition model.

[0053] In an optional implementation, the first acquiring unit includes:

[0054] The first acquisition subunit is used to acquire a sentiment classification acquisition model suitable for the user from a plurality of trained sentiment classification acquisition models according to the historical manipulation behavior data.

[0055] In an optional implementation, the first acquisition subunit is specifically used to: determine the user set to which the user belongs from multiple user sets based on the historical manipulation behavior data; and select the emotion classification acquisition model applicable to the user set to which the user belongs from the emotion classification acquisition models applicable to each user set respectively.

[0056] In an optional implementation, the third acquisition module includes:

[0057] A first determination unit is used to determine, for any one manipulation type of manipulation behavior supported by the current object, the current execution status of the manipulation behavior of the current object with respect to the manipulation type by the user according to the current manipulation behavior data, and a second determination unit is used to determine, according to the historical manipulation behavior data, the historical execution status of the manipulation behavior of the historical object with respect to the manipulation type by the user, and a third acquisition unit is used to acquire the emotion score of the user with respect to the manipulation behavior of the current object with respect to the manipulation type according to the current execution status and the historical execution status;

[0058] The fourth acquisition unit is used to acquire the emotion classification of the user towards the current object according to the emotion scores of the user's manipulation behaviors on the current object regarding each manipulation type.

[0059] In an optional implementation, the third acquisition unit includes:

[0060] A second acquisition subunit is used to acquire a basic emotion classification of the manipulation behavior of the manipulation type;

[0061] a statistical subunit, for collecting statistical results of whether the user has an execution tendency to perform a manipulation behavior of the manipulation type on the historical object according to the historical execution situation, and a determining subunit, for determining an execution result of whether the user has performed a manipulation behavior of the manipulation type on the current object according to the current execution situation;

[0062] The third acquisition subunit is used to obtain the user's emotion score for the manipulation behavior of the current object regarding the manipulation type according to the basic emotion classification of the manipulation behavior of the manipulation type, the statistical result and the execution result.

[0063] In an optional implementation, the statistical subunit is specifically used to: obtain the historical number of historical objects that the user has manipulated in the historical process, and obtain the number of times the user has executed the manipulation type of manipulation on the historical object in the process of manipulating the historical object in the historical process; and obtain statistical results of whether the user has a tendency to execute the manipulation type of manipulation on the historical object based on the historical number and the number of executions.

[0064] In an optional implementation, the statistical subunit is specifically used to: calculate the ratio between the number of executions and the historical number; when the ratio is greater than or equal to a preset ratio, determine that the user has an execution tendency to perform a manipulation behavior of the manipulation type on the historical object; or, when the ratio is less than a preset ratio, determine that the user does not have an execution tendency to perform a manipulation behavior of the manipulation type on the historical object.

[0065] In an optional implementation, the third acquisition subunit is specifically used to: in the correspondence between the basic emotion classification of the manipulation behavior, the statistical results of the execution tendency of the manipulation behavior, the execution results of the manipulation behavior and the user's emotion score regarding the manipulation behavior, search for the user's emotion score regarding the manipulation behavior that corresponds to the basic emotion classification of the manipulation behavior of the manipulation type, the statistical results and the execution results, and use it as the user's emotion score for the manipulation behavior of the current object regarding the manipulation type.

[0066] In an optional implementation, the fourth obtaining unit includes:

[0067] A summing subunit, configured to weight and sum the emotion scores of the user's manipulation behaviors on the current object with respect to each manipulation type, to obtain a total emotion score of the user on the current object;

[0068] The fourth acquisition subunit is used to acquire the emotion classification of the user towards the current object according to the total emotion value of the user towards the current object.

[0069] In an optional implementation, the fourth acquisition subunit is specifically used to: when the total emotion value of the user towards the current object is greater than a preset emotion threshold, determine that the user's emotion towards the current object is classified as a positive emotion; or, when the total emotion value of the user towards the current object is equal to the preset emotion threshold, determine that the user's emotion towards the current object is classified as a neutral emotion; or, when the total emotion value of the user towards the current object is less than the preset emotion threshold, determine that the user's emotion towards the current object is classified as a negative emotion.

[0070] In a third aspect, the present application shows an electronic device, the electronic device comprising:

[0071] processor;

[0072] a memory for storing processor-executable instructions;

[0073] Wherein, the processor is configured to execute the data processing method as described in the first aspect.

[0074] In a fourth aspect, the present application illustrates a non-temporary computer-readable storage medium, which, when instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to perform the data processing method as described in the first aspect.

[0075] In a fifth aspect, the present application illustrates a computer program product. When instructions in the computer program product are executed by a processor of an electronic device, the electronic device is enabled to perform the data processing method as described in the first aspect.

[0076] The technical solution provided by this application may have the following beneficial effects:

[0077] After analysis by the inventor, it is concluded that: judging the user's emotional classification of the current object only based on the user's manipulation behavior on the current object, that is, using the user's manipulation behavior on the current object to absolutely judge the user's emotional classification of the current object, may very likely lead to inaccurate emotional classification of the user on the current object.

[0078] For example, the basic emotion classification of the manipulation behavior of "clicking and disliking" is negative emotion (if the user clicks and dislikes an object, it means that the user does not like this object. If the user does not click and dislike this object, it does not mean that the user likes this object or dislikes this object. It is often impossible to determine whether the user likes this object or dislikes this object based on the state of "the user did not click and dislike this object", so the user's emotion classification for this object is often determined as neutral emotion). Among them, when the user browses short videos one by one using a short video application, if he clicks and dislikes almost every short video he browses, clicking and disliking a short video often means that he does not like the short video. If the user does not click and dislike this short video when browsing a short video later, combined with the user's habits, the user often likes this short video (because according to the user's habits, if the user does not like it, he will click and dislike it). "Not clicking and disliking this short video" can be regarded as an implicit feedback of positive emotion. However, simply based on the fact that "the user did not like the short video", the user's emotion classification for the short video is often determined as neutral, which is inconsistent with the user's positive emotion that he or she likes the short video, resulting in inaccurate emotion classification of the user for the current object.

[0079] For another example, the basic emotion classification of the manipulation behavior of "like" is positive emotion (if the user likes an object, it means that the user likes this object. If the user does not like this object, it does not mean that the user does not like this object or like this object. It is often impossible to determine whether the user likes or dislikes this object based on the state of "the user did not like this object", so the user's emotion classification for this object is often determined to be neutral emotion). Among them, when the user browses short videos one by one using a short video application, if he likes almost every short video he browses, liking a short video often means that he likes the short video. If the user does not like a short video when browsing a short video later, combined with the user's habits, the user often does not like the short video (because according to the user's habits, if the user likes it, he will like it). "Not liking the short video" can be regarded as an implicit feedback of negative emotion. However, simply based on the fact that "the user did not like the short video", the user's emotion classification for the short video is often determined as neutral, which is inconsistent with the user's positive emotion that he likes the short video, resulting in inaccurate emotion classification of the user for the current object.

[0080] Secondly, the historical manipulation behavior data of the user on the historical objects in the historical process can reflect the historical behavior habits of the user in the historical process, and the current manipulation behavior data of the user on the current object can reflect the current manipulation state of the user.

[0081] If the current manipulation state is significantly different from the user's historical behavior habits in the historical process, the user's current manipulation behavior data on the current object can often reflect the user's emotional classification of the current object as negative emotion or positive emotion.

[0082] In view of this, in the present application, when the user manipulates the current object, for each manipulation type supported by the current object, the user's current manipulation behavior data on the current object is obtained, and the user's current manipulation behavior data on the current object includes the manipulation type of the manipulation behavior actually performed by the user on the current object and the manipulation type of the manipulation behavior not performed. For each historical object that the user has manipulated in the historical process, for each manipulation type supported by the historical object, the user's historical manipulation behavior data on the historical object is obtained, and the user's historical manipulation behavior data on the historical object includes the manipulation type of the manipulation behavior actually performed by the user on the historical object and the manipulation type of the manipulation behavior not performed. Based on the user's current manipulation behavior data on the current object and the user's historical manipulation behavior data on the historical objects, the user's emotional classification of the current object is obtained.

[0083] Through this application, the user's historical manipulation behavior data on historical objects in the historical process is used as a benchmark, that is, the user's historical behavioral habits in the historical process are used as a benchmark, and the user's current manipulation behavior data on the current object are combined to obtain the user's emotional classification of the current object. Therefore, the user's emotional classification of the current object can be obtained according to the user's actual situation (such as behavioral habits, etc.), thereby improving the accuracy of the obtained user's emotional classification of the current object. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] Figure 1 It is a step flow chart of a data processing method of the present application.

[0085] Figure 2 It is a step flow chart of a data processing method of the present application.

[0086] Figure 3 It is a step flow chart of a data processing method of the present application.

[0087] Figure 4 It is a step flow chart of a data processing method of the present application.

[0088] Figure 5 It is a structural block diagram of a data processing device of the present application.

[0089] Figure 6 It is a block diagram of an electronic device of the present application.

[0090] Figure 7 It is a block diagram of an electronic device of the present application. DETAILED DESCRIPTION

[0091] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0092] Reference Figure 1 , shows a flow chart of the steps of a data processing method of the present application, the method is applied to a terminal, and the method may specifically include the following steps:

[0093] In step S101, when the user manipulates the current object, the user's current manipulation behavior data on the current object is obtained for each manipulation type supported by the current object. The user's current manipulation behavior data on the current object includes the manipulation types of manipulation behaviors actually performed by the user on the current object and the manipulation types of manipulation behaviors not performed.

[0094] The manipulation behaviors supported by the current object include: browsing with multiple different durations (such as short-duration browsing, medium-duration browsing, and long-duration browsing, etc., and the multiple different durations can be determined according to actual conditions, such as 0 to 20 seconds, 20 seconds to 40 seconds, and 40 seconds to 60 seconds, etc.), likes, dislikes, reports, complaints, shares, favorites, adding to shopping carts, purchases, registrations, downloads, and forwarding, etc. These manipulation behaviors are all manipulation behaviors of different types.

[0095] The current manipulation behavior data of the user on the current object includes: whether the user has performed manipulation behaviors of various manipulation types supported by the current object.

[0096] For example, the user performed one of the browsing behaviors for a continuous period of time on the current object, did not perform the like behavior, did not perform the dislike behavior, did not perform the report behavior, did not perform the complaint behavior, did not perform the share behavior, did not perform the favorite behavior, did not perform the add to shopping cart behavior, did not perform the purchase behavior, did not perform the registration behavior, did not perform the download behavior, and did not perform the forwarding behavior, etc.

[0097] The current manipulation behavior data of the user on the current object may be counted in real time when the user manipulates the current object.

[0098] Among them, in the process of the user browsing the current object, among the manipulation behaviors of various manipulation types supported by the current object, if the user does not perform a manipulation behavior of a certain manipulation type on the current object, then "not performing a manipulation behavior of a certain manipulation type" can be regarded as the user's implicit feedback on the current object.

[0099] Alternatively, during the process of the user browsing the current object, the manipulation behavior of the manipulation type not performed by the user on the current object can be regarded as the user's implicit feedback on the current object.

[0100] In step S102, for each historical object that the user has manipulated in the historical process, the user's historical manipulation behavior data on the historical object is obtained for each manipulation type supported by the historical object. The user's historical manipulation behavior data on the historical object includes the manipulation types of the manipulation behaviors actually performed by the user on the historical object and the manipulation types of the manipulation behaviors not performed.

[0101] Whenever a user manipulates a historical object in the historical process, the historical manipulation behavior data of the user on the historical object will be automatically recorded. Therefore, the historical manipulation behavior data of the user on the historical object can be obtained according to the recorded content.

[0102] The various historical objects that the user has manipulated in the historical process may include: various historical objects that the user has manipulated in a period close to the current moment, etc., to improve timeliness.

[0103] The manipulation behaviors supported by the history object include: browsing with different durations (such as short-duration browsing, medium-duration browsing, and long-duration browsing, etc., and the different durations can be determined according to actual conditions, such as 0 to 20 seconds, 20 seconds to 40 seconds, and 40 seconds to 60 seconds, etc.), likes, dislikes, reports, complaints, shares, favorites, adding to shopping carts, purchases, registrations, downloads, and forwarding, etc. These manipulation behaviors are all manipulation behaviors of different types.

[0104] The historical manipulation behavior data of the user on the historical object includes: whether the user has performed manipulation behaviors on each manipulation type supported by the historical object.

[0105] For example, the user performed one of the browsing behaviors for a certain duration on the historical object, did not perform the like behavior, did not perform the dislike behavior, did not perform the report behavior, did not perform the complaint behavior, did not perform the share behavior, did not perform the favorite behavior, did not perform the add to shopping cart behavior, did not perform the purchase behavior, did not perform the registration behavior, did not perform the download behavior, and did not perform the forwarding behavior, etc.

[0106] In step S103, the user's emotion classification on the current object is obtained according to the user's current manipulation behavior data on the current object and the user's historical manipulation behavior data on historical objects.

[0107] Sentiment classification includes positive sentiment, neutral sentiment, and negative sentiment.

[0108] Among them, when the user's emotion classification for the current object is positive, the user often likes the current object. When the user's emotion classification for the current object is negative, the user often dislikes the current object. When the user's emotion classification for the current object is neutral, the user often does not clearly like or dislike the current object.

[0109] Of course, we can further divide the sentiment into more fine-grained categories.

[0110] For example, positive emotions are divided into multiple different levels, and the degree of liking reflected by high-level positive emotions is higher than the degree of liking reflected by low-level positive emotions.

[0111] For another example, negative emotions are divided into multiple different levels, and the degree of disgust embodied by high-level negative emotions is higher than the degree of disgust embodied by low-level negative emotions.

[0112] The details of this step can be found in the following Figure 2 or Figure 3 The illustrated embodiment will not be described in detail here.

[0113] After analysis by the inventor, it is concluded that: judging the user's emotional classification of the current object only based on the user's manipulation behavior on the current object, that is, using the user's manipulation behavior on the current object to absolutely judge the user's emotional classification of the current object, may very likely lead to inaccurate emotional classification of the user on the current object.

[0114] For example, the basic emotion classification of the manipulation behavior of "clicking and disliking" is negative emotion (if the user clicks and dislikes an object, it means that the user does not like this object. If the user does not click and dislike this object, it does not mean that the user likes this object or dislikes this object. It is often impossible to determine whether the user likes this object or dislikes this object based on the state of "the user did not click and dislike this object", so the user's emotion classification for this object is often determined as neutral emotion). Among them, when the user browses short videos one by one using a short video application, if he clicks and dislikes almost every short video he browses, clicking and disliking a short video often means that he does not like the short video. If the user does not click and dislike this short video when browsing a short video later, combined with the user's habits, the user often likes this short video (because according to the user's habits, if the user does not like it, he will click and dislike it). "Not clicking and disliking this short video" can be regarded as an implicit feedback of positive emotion. However, simply based on the fact that "the user did not like the short video", the user's emotion classification for the short video is often determined as neutral, which is inconsistent with the user's positive emotion that he or she likes the short video, resulting in inaccurate emotion classification of the user for the current object.

[0115] For another example, the basic emotion classification of the manipulation behavior of "like" is positive emotion (if the user likes an object, it means that the user likes this object. If the user does not like this object, it does not mean that the user does not like this object or like this object. It is often impossible to determine whether the user likes or dislikes this object based on the state of "the user did not like this object", so the user's emotion classification for this object is often determined to be neutral emotion). Among them, when the user browses short videos one by one using a short video application, if he likes almost every short video he browses, liking a short video often means that he likes the short video. If the user does not like a short video when browsing a short video later, combined with the user's habits, the user often does not like the short video (because according to the user's habits, if the user likes it, he will like it). "Not liking the short video" can be regarded as an implicit feedback of negative emotion. However, simply based on the fact that "the user did not like the short video", the user's emotion classification for the short video is often determined as neutral, which is inconsistent with the user's positive emotion that he likes the short video, resulting in inaccurate emotion classification of the user for the current object.

[0116] Secondly, the historical manipulation behavior data of the user on the historical objects in the historical process can reflect the historical behavior habits of the user in the historical process, and the current manipulation behavior data of the user on the current object can reflect the current manipulation state of the user.

[0117] If the current manipulation state is significantly different from the user's historical behavior habits in the historical process, the user's current manipulation behavior data on the current object can often reflect the user's emotional classification of the current object as negative emotion or positive emotion.

[0118] In view of this, in the present application, when the user manipulates the current object, for each manipulation type supported by the current object, the user's current manipulation behavior data on the current object is obtained, and the user's current manipulation behavior data on the current object includes the manipulation type of the manipulation behavior actually performed by the user on the current object and the manipulation type of the manipulation behavior not performed. For each historical object that the user has manipulated in the historical process, for each manipulation type supported by the historical object, the user's historical manipulation behavior data on the historical object is obtained, and the user's historical manipulation behavior data on the historical object includes the manipulation type of the manipulation behavior actually performed by the user on the historical object and the manipulation type of the manipulation behavior not performed. Based on the user's current manipulation behavior data on the current object and the user's historical manipulation behavior data on the historical objects, the user's emotional classification of the current object is obtained.

[0119] Through this application, the user's historical manipulation behavior data on historical objects in the historical process is used as a benchmark, that is, the user's historical behavioral habits in the historical process are used as a benchmark, and the user's current manipulation behavior data on the current object are combined to obtain the user's emotional classification of the current object. Therefore, the user's emotional classification of the current object can be obtained according to the user's actual situation (such as behavioral habits, etc.), thereby improving the accuracy of the obtained user's emotional classification of the current object.

[0120] In one embodiment of the present application, see Figure 2 , step S103 includes:

[0121] In step S201, a sentiment classification acquisition model is obtained, and the sentiment classification acquisition model is trained based on multiple training data sets, and the training data sets include: first sample manipulation behavior data of a sample user on a first sample object, second sample manipulation behavior data of a sample user on a second sample object, and labeled sentiment classification of the sample user on the second sample object.

[0122] Among them, the sentiment classification acquisition model can be trained in advance.

[0123] For example, multiple training data sets may be obtained, the training data sets including: first sample manipulation behavior data of a sample user on a first sample object, second sample manipulation behavior data of a sample user on a second sample object, and labeled emotion classification of the sample user on the second sample object.

[0124] The first sample manipulation behavior data of the sample user on the first sample object includes: the manipulation type of the manipulation behavior actually performed by the sample user on the first sample object and the manipulation type of the manipulation behavior not performed.

[0125] The second sample manipulation behavior data of the sample user on the second sample object includes: the manipulation type of the manipulation behavior actually performed by the sample user on the second sample object and the manipulation type of the manipulation behavior not performed.

[0126] Then, the model can be trained using multiple training data sets until the parameters in the model converge, thereby obtaining a sentiment classification acquisition model.

[0127] Models include convolutional neural networks or recurrent neural networks, etc.

[0128] The trained sentiment classification acquisition model can be deployed in the terminal, so the terminal can directly obtain the sentiment classification acquisition model deployed in the terminal, and then step S202 can be executed.

[0129] In step S202, the user's current manipulation behavior data on the current object and the user's historical manipulation behavior data on historical objects are input into the emotion classification acquisition model, so that the emotion classification acquisition model processes the user's current manipulation behavior data on the current object and the user's historical manipulation behavior data on historical objects, obtains the user's emotion classification of the current object, and outputs the user's emotion classification of the current object.

[0130] In step S203, the sentiment classification of the user on the current object output by the sentiment classification acquisition model is obtained.

[0131] Furthermore, in one embodiment of the present application, multiple emotion classification acquisition models may be trained in advance, one emotion classification acquisition model is applicable to a portion of users, and users applicable to different emotion classification acquisition models may not overlap.

[0132] In this way, when acquiring the emotion classification acquisition model, an emotion classification acquisition model suitable for the user can be acquired from multiple trained emotion classification acquisition models based on the user's historical manipulation behavior data on historical objects.

[0133] In this way, when obtaining the user's emotion classification of the current object based on the user's current manipulation behavior data on the current object and the user's historical manipulation behavior data on historical objects, a emotion classification acquisition model suitable for the user can be used instead of a public emotion classification acquisition model, thereby further improving the accuracy of obtaining the user's emotion classification of the current object.

[0134] Among them, when obtaining a sentiment classification acquisition model suitable for the user from multiple trained sentiment classification acquisition models based on the user's historical manipulation behavior data on historical objects, the user set to which the user belongs can be determined from multiple user sets based on the user's historical manipulation behavior data on historical objects, and then the sentiment classification acquisition model suitable for the user set to which the user belongs can be selected from the sentiment classification acquisition models respectively suitable for each user set.

[0135] Among them, different user sets are suitable for different sentiment classification acquisition models.

[0136] The historical manipulation behavior data of the users in one user set on the historical object are similar, while the historical manipulation behavior data of the users in different user sets on the historical object are not similar.

[0137] In this way, the user set to which the user belongs can be determined from multiple user sets according to the historical manipulation behavior data of the user on the historical objects in the following manner.

[0138] For example, for any user set, obtain the statistical data of the historical manipulation behavior data of each user in the user set on the historical object. For example, obtain the number of times each user in the user set performs each manipulation type of manipulation operation on the historical object, then sum up the number of times the manipulation operations of each manipulation type are executed to obtain the total number of executions, calculate the ratio between the number of times the manipulation operations of each manipulation type are executed and the total number of executions, and obtain the execution ratio of the manipulation operations of each manipulation type in the user set. The statistical data can reflect the manipulation characteristics (historical behavior habits, etc.) of each user in the user set on the historical object.

[0139] The same goes for each of the other user sets.

[0140] Then, the statistical data of the historical manipulation behavior data of the user on the historical object can be obtained. For example, the number of times the user performs each type of manipulation operation on the historical object is obtained, and then the number of times the manipulation operations of each type of manipulation are performed is summed up to obtain the total number of executions, and the ratio between the number of times the manipulation operations of each type of manipulation are performed and the total number of executions is calculated to obtain the execution ratio of the manipulation operations of each type of manipulation by the user. The statistical data can reflect the manipulation characteristics of the user on the historical object (historical behavior habits, etc.).

[0141] After that, the similarity between the user's statistical data and the statistical data of each user set can be calculated. For example, for any user set, for any manipulation type, the square of the difference between the execution ratio of the manipulation operation of the manipulation type of the user and the execution ratio of the manipulation operation of the manipulation type of the user set can be calculated to obtain the square corresponding to the manipulation type, and the same is true for each other manipulation type. Then, the square root of the sum of the squares corresponding to each manipulation type is calculated as the similarity between the user's statistical data and the statistical data of the user set. The same is true for each other user set.

[0142] Then, the user set with the highest similarity between the statistical data and the statistical data of the user can be used as the user set to which the user belongs.

[0143] In another embodiment of the present application, a training method for acquiring a sentiment classification model applicable to different user sets includes:

[0144] For each manipulation behavior of each manipulation type supported by the first sample object, first sample manipulation behavior data of each sample user on the first sample object is obtained. The first sample manipulation behavior data includes the manipulation type of the manipulation behavior actually performed by the sample user on the first sample object and the manipulation type of the manipulation behavior not performed.

[0145] According to the first sample manipulation behavior data of each sample user on the first sample object, the multiple sample users are classified to obtain at least two sample user sets.

[0146] For example, multiple sample users may be classified according to the first sample manipulation behavior data of each sample user on the first sample object in an existing manner (such as K-nearest neighbor algorithm, etc.), and sample users in different sample user sets do not overlap.

[0147] The historical manipulation behavior data of each user in a sample user set on the historical object are similar, while the historical manipulation behavior data of users in different sample user sets on the historical object are not similar.

[0148] For any sample user set, the sentiment classification acquisition model applicable to the sample user set can be obtained in the following way, and the same is true for each other sample user set.

[0149] For any first sample user in the sample user set, for each manipulation behavior of the manipulation type supported by the second sample object, obtain the second sample manipulation behavior data of the first sample user on the second sample object, the second sample manipulation behavior data includes the manipulation type of the manipulation behavior actually performed by the first sample user on the second sample object and the manipulation type of the manipulation behavior not performed, and obtain the labeled emotion classification of the first sample user on the second sample object. The first sample manipulation behavior data of the first sample user on the first sample object, the second sample manipulation behavior data of the first sample user on the second sample object, and the labeled emotion classification of the first sample user on the second sample object are combined into the training data set of the first sample user. For each sample user in the sample user set, the same is done to obtain the training data set of each sample user in the sample user set.

[0150] The model is trained according to the training data sets of each first sample user in the sample user set until the parameters in the model converge, thereby obtaining a sentiment classification acquisition model suitable for the sample user set.

[0151] Among them, the models include convolutional neural networks or recurrent neural networks, etc.

[0152] In one embodiment of the present application, see Figure 3 , step S103 includes:

[0153] In step S301, for any manipulation behavior of a manipulation type supported by the current object, the current execution status of the user's manipulation behavior of the current object regarding the manipulation type is determined based on the user's current manipulation behavior data of the current object, and the historical execution status of the user's manipulation behavior of the historical object regarding the manipulation type is determined based on the user's historical manipulation behavior data of the historical object, and the user's emotional score for the manipulation behavior of the current object regarding the manipulation type is obtained based on the current execution status and the historical execution status.

[0154] The current execution status of the manipulation behavior of the user on the current object with respect to the manipulation type includes: whether the user has executed the manipulation behavior with respect to the manipulation type on the current object or has not executed the manipulation behavior with respect to the manipulation type.

[0155] The historical execution status of the user's manipulation behavior on the historical object with respect to the manipulation type includes: whether the user has executed the manipulation behavior with respect to the manipulation type on the historical object or has not executed the manipulation behavior with respect to the manipulation type.

[0156] The specific method of obtaining the user's emotional score for the manipulation behavior of the current object on the manipulation type according to the current execution status and the historical execution status can be found in the following example. Figure 4 The illustrated embodiment will not be described in detail here.

[0157] In step S302, the user's emotion classification for the current object is obtained according to the emotion scores of the user's manipulation behaviors for each manipulation type of the current object.

[0158] In this step, the emotional scores of the user's manipulation behaviors for each manipulation type on the current object can be weighted and summed to obtain the user's total emotional value for the current object, and then the user's emotional classification for the current object can be obtained based on the user's total emotional value for the current object.

[0159] For example, when the total value of the user's emotion towards the current object is greater than the preset emotion threshold, the user's emotion towards the current object is determined to be classified as positive emotion. Alternatively, when the total value of the user's emotion towards the current object is equal to the preset emotion threshold, the user's emotion towards the current object is determined to be classified as neutral emotion. Alternatively, when the total value of the user's emotion towards the current object is less than the preset emotion threshold, the user's emotion towards the current object is determined to be classified as negative emotion.

[0160] Among them, the preset emotion threshold can be determined according to actual conditions, and this application does not impose any limitation on this.

[0161] In one embodiment of the present application, see Figure 4 , step S301 includes:

[0162] In step S401, a basic emotion classification of the manipulation behavior of the manipulation type is obtained.

[0163] In the present application, each type of manipulation behavior in the manipulation behaviors supported by the current object has its own basic emotion classification, and the basic emotion classification includes positive emotion and negative emotion, etc.

[0164] For example, the basic emotion classification of likes is positive emotion, the basic emotion classification of dislikes is negative emotion, the basic emotion classification of reports is negative emotion, the basic emotion classification of complaints is negative emotion, the basic emotion classification of shares is positive emotion, the basic emotion classification of favorites is positive emotion, the basic emotion classification of adding to shopping carts is positive emotion, the basic emotion classification of purchases is positive emotion, the basic emotion classification of registrations is positive emotion, the basic emotion classification of downloads is positive emotion, and the basic emotion classification of forwarding is positive emotion, etc.

[0165] The basic emotion classification of the manipulation behavior of each manipulation type in the manipulation behaviors supported by the current object may be manually set in advance based on experience and actual conditions.

[0166] In step S402, statistics are collected based on historical execution conditions to determine whether the user has a tendency to perform a manipulation behavior of the manipulation type on the historical object, and based on the current execution conditions, the execution result of determining whether the user has performed a manipulation behavior of the manipulation type on the current object is determined.

[0167] In particular, when the statistical result of whether the user has the execution tendency of executing the manipulation behavior of the manipulation type on the historical object is counted according to the historical execution situation, it can be achieved through the following process, including:

[0168] 4021. Obtain the historical number of historical objects that the user has manipulated in the historical process, and obtain the number of times the user has performed a manipulation action of the manipulation type on the historical object in the process of manipulating the historical object in the historical process.

[0169] 4021. Statistical results of whether the user has an execution tendency to execute the manipulation behavior of the manipulation type on the historical object are counted based on the history quantity and the execution times.

[0170] For example, the ratio between the number of executions and the number of history records may be calculated. When the ratio is greater than or equal to a preset ratio, it is determined that the user has an execution tendency to perform a manipulation behavior of the manipulation type on the history object. Alternatively, when the ratio is less than a preset ratio, it is determined that the user does not have an execution tendency to perform a manipulation behavior of the manipulation type on the history object.

[0171] The preset ratio can be determined according to actual conditions, and this application does not impose any limitation on this.

[0172] In step S403, the user's emotion score for the manipulation behavior of the current object with respect to the manipulation type is obtained according to the basic emotion classification, statistical results, and execution results of the manipulation behavior of the manipulation type.

[0173] In one example, when the basic emotion classification of the manipulation behavior of the manipulation type is positive emotion, if the statistical result is used to indicate that the user has an execution tendency to perform the manipulation behavior of the manipulation type on the historical object, and the execution result is used to indicate that the user has not performed the manipulation behavior of the manipulation type on the current object, then the user's emotion score for the manipulation behavior of the manipulation type on the current object is set to the first negative emotion score.

[0174] Alternatively, when the basic emotion classification of the manipulation behavior of the manipulation type is positive emotion, if the statistical result is used to indicate that the user has an execution tendency to perform the manipulation behavior of the manipulation type on the historical object, and the execution result is used to indicate that the user has performed the manipulation behavior of the manipulation type on the current object, then the user's emotion score for the manipulation behavior of the manipulation type on the current object is set to a neutral emotion score. The neutral emotion score is greater than the first negative emotion score.

[0175] Alternatively, when the basic emotion classification of the manipulation behavior of this manipulation type is positive emotion, if the statistical result is used to indicate that the user does not have the execution tendency to perform the manipulation behavior of this manipulation type on the historical object, and the execution result is used to indicate that the user has not performed the manipulation behavior of this manipulation type on the current object, then the user's emotion score for the manipulation behavior of this manipulation type on the current object is set to a neutral emotion score.

[0176] Alternatively, when the basic emotion classification of the manipulation behavior of the manipulation type is positive emotion, if the statistical result is used to indicate that the user does not have an execution tendency to perform the manipulation behavior of the manipulation type on the historical object, and the execution result is used to indicate that the user has performed the manipulation behavior of the manipulation type on the current object, the emotion score of the user on the manipulation behavior of the manipulation type on the current object is set to the first positive emotion score. The first positive emotion score is greater than the neutral emotion score.

[0177] Alternatively, when the basic emotion classification of the manipulation behavior of this manipulation type is negative emotion, if the statistical result is used to indicate that the user has an execution tendency to perform the manipulation behavior of this manipulation type on the historical object, and the execution result is used to indicate that the user has not performed the manipulation behavior of this manipulation type on the current object, then the user's emotion score for the manipulation behavior of this manipulation type on the current object is set to the second positive emotion score.

[0178] Alternatively, when the basic emotion classification of the manipulation behavior of the manipulation type is negative, if the statistical result is used to indicate that the user has an execution tendency to perform the manipulation behavior of the manipulation type on the historical object, and the execution result is used to indicate that the user has performed the manipulation behavior of the manipulation type on the current object, then the user's emotion score for the manipulation behavior of the manipulation type on the current object is set to a neutral emotion score. The neutral emotion score is less than the second positive emotion score.

[0179] Alternatively, when the basic emotion classification of the manipulation behavior of this manipulation type is negative emotion, if the statistical result is used to indicate that the user does not have the execution tendency to perform the manipulation behavior of this manipulation type on the historical object, and the execution result is used to indicate that the user has not performed the manipulation behavior of this manipulation type on the current object, then the user's emotion score for the manipulation behavior of this manipulation type on the current object is set to a neutral emotion score.

[0180] Alternatively, when the basic emotion classification of the manipulation behavior of the manipulation type is negative, if the statistical result is used to indicate that the user does not have an execution tendency to perform the manipulation behavior of the manipulation type on the historical object, and the execution result is used to indicate that the user has performed the manipulation behavior of the manipulation type on the current object, the user's emotion score for the manipulation behavior of the manipulation type on the current object is set to a second negative emotion score. The second negative emotion score is less than the neutral emotion score.

[0181] In this way, the correspondence between the basic emotion classification of the manipulation behavior, the statistical results of the execution tendency of the manipulation behavior, the execution result of the manipulation behavior and the user's emotion score for the manipulation behavior can be set in advance. The correspondence includes corresponding table items of different basic emotion classifications of manipulation behaviors, different statistical results of the execution tendency of the manipulation behavior, different execution results of the manipulation behavior and the user's emotion score for different manipulation behaviors. In this way, in the correspondence between the basic emotion classification of the manipulation behavior, the statistical results of the execution tendency of the manipulation behavior, the execution result of the manipulation behavior and the user's emotion score for the manipulation behavior, the user's emotion score for the manipulation behavior that corresponds to the basic emotion classification, statistical results and execution result of the manipulation behavior of the manipulation type can be found and used as the user's emotion score for the manipulation behavior of the current object regarding the manipulation type.

[0182] It should be noted that, for the method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the order of the actions described, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions involved are not necessarily required by the present application.

[0183] Reference Figure 5 , shows a structural block diagram of a data processing device of the present application, which is applied to a terminal, and the device includes:

[0184] A first acquisition module 11 is used to acquire, when a user manipulates a current object, current manipulation behavior data of the user on the current object for each manipulation type supported by the current object, wherein the current manipulation behavior data includes the manipulation type of the manipulation behavior actually performed by the user on the current object and the manipulation type of the manipulation behavior not performed;

[0185] A second acquisition module 12 is used to acquire, for each historical object that the user has manipulated in the historical process, historical manipulation behavior data of the user on the historical object for each manipulation type supported by the historical object, wherein the historical manipulation behavior data includes the manipulation type of the manipulation behavior actually performed by the user on the historical object and the manipulation type of the manipulation behavior not performed;

[0186] The third acquisition module 13 is used to acquire the user's emotion classification of the current object according to the current manipulation behavior data and the historical manipulation behavior data.

[0187] In an optional implementation, the third acquisition module includes:

[0188] A first acquisition unit is used to acquire a sentiment classification acquisition model, wherein the sentiment classification acquisition model is obtained by training based on a plurality of training data sets, wherein the training data sets include: first sample manipulation behavior data of a sample user on a first sample object, second sample manipulation behavior data of the sample user on a second sample object, and a labeled sentiment classification of the sample user on the second sample object;

[0189] An input unit, used to input the current manipulation behavior data and the historical manipulation behavior data into the emotion classification acquisition model, so that the emotion classification acquisition model processes the current manipulation behavior data and the historical manipulation behavior data to obtain the user's emotion classification of the current object, and output the user's emotion classification of the current object;

[0190] The second acquisition unit is used to acquire the sentiment classification of the user to the current object output by the sentiment classification acquisition model.

[0191] In an optional implementation, the first acquiring unit includes:

[0192] The first acquisition subunit is used to acquire a sentiment classification acquisition model suitable for the user from a plurality of trained sentiment classification acquisition models according to the historical manipulation behavior data.

[0193] In an optional implementation, the first acquisition subunit is specifically used to: determine the user set to which the user belongs from multiple user sets based on the historical manipulation behavior data; and select the emotion classification acquisition model applicable to the user set to which the user belongs from the emotion classification acquisition models applicable to each user set respectively.

[0194] In an optional implementation, the third acquisition module includes:

[0195] A first determination unit is used to determine, for any one manipulation type of manipulation behavior supported by the current object, the current execution status of the manipulation behavior of the current object with respect to the manipulation type by the user according to the current manipulation behavior data, and a second determination unit is used to determine, according to the historical manipulation behavior data, the historical execution status of the manipulation behavior of the historical object with respect to the manipulation type by the user, and a third acquisition unit is used to acquire the emotion score of the user with respect to the manipulation behavior of the current object with respect to the manipulation type according to the current execution status and the historical execution status;

[0196] The fourth acquisition unit is used to acquire the emotion classification of the user towards the current object according to the emotion scores of the user's manipulation behaviors of each manipulation type towards the current object.

[0197] In an optional implementation, the third acquisition unit includes:

[0198] A second acquisition subunit is used to acquire a basic emotion classification of the manipulation behavior of the manipulation type;

[0199] a statistical subunit, for collecting statistical results of whether the user has an execution tendency to perform a manipulation behavior of the manipulation type on the historical object according to the historical execution situation, and a determining subunit, for determining an execution result of whether the user has performed a manipulation behavior of the manipulation type on the current object according to the current execution situation;

[0200] The third acquisition subunit is used to obtain the user's emotion score for the manipulation behavior of the current object regarding the manipulation type according to the basic emotion classification of the manipulation behavior of the manipulation type, the statistical result and the execution result.

[0201] In an optional implementation, the statistical subunit is specifically used to: obtain the historical number of historical objects that the user has manipulated in the historical process, and obtain the number of times the user has executed the manipulation type of manipulation on the historical object in the process of manipulating the historical object in the historical process; and obtain statistical results of whether the user has a tendency to execute the manipulation type of manipulation on the historical object based on the historical number and the number of executions.

[0202] In an optional implementation, the statistical subunit is specifically used to: calculate the ratio between the number of executions and the historical number; when the ratio is greater than or equal to a preset ratio, determine that the user has an execution tendency to perform a manipulation behavior of the manipulation type on the historical object; or, when the ratio is less than a preset ratio, determine that the user does not have an execution tendency to perform a manipulation behavior of the manipulation type on the historical object.

[0203] In an optional implementation, the third acquisition subunit is specifically used to: in the correspondence between the basic emotion classification of the manipulation behavior, the statistical results of the execution tendency of the manipulation behavior, the execution results of the manipulation behavior and the user's emotion score regarding the manipulation behavior, search for the user's emotion score regarding the manipulation behavior that corresponds to the basic emotion classification of the manipulation behavior of the manipulation type, the statistical results and the execution results, and use it as the user's emotion score for the manipulation behavior of the current object regarding the manipulation type.

[0204] In an optional implementation, the fourth obtaining unit includes:

[0205] A summing subunit, configured to weight and sum the emotion scores of the user's manipulation behaviors on the current object with respect to each manipulation type, to obtain a total emotion score of the user on the current object;

[0206] The fourth acquisition subunit is used to acquire the emotion classification of the user towards the current object according to the total emotion value of the user towards the current object.

[0207] In an optional implementation, the fourth acquisition subunit is specifically used to: when the total emotion value of the user towards the current object is greater than a preset emotion threshold, determine that the user's emotion towards the current object is classified as a positive emotion; or, when the total emotion value of the user towards the current object is equal to the preset emotion threshold, determine that the user's emotion towards the current object is classified as a neutral emotion; or, when the total emotion value of the user towards the current object is less than the preset emotion threshold, determine that the user's emotion towards the current object is classified as a negative emotion.

[0208] After analysis by the inventor, it is concluded that: judging the user's emotional classification of the current object only based on the user's manipulation behavior on the current object, that is, using the user's manipulation behavior on the current object to absolutely judge the user's emotional classification of the current object, may very likely lead to inaccurate emotional classification of the user on the current object.

[0209] For example, the basic emotion classification of the manipulation behavior of "clicking and disliking" is negative emotion (if the user clicks and dislikes an object, it means that the user does not like this object. If the user does not click and dislike this object, it does not mean that the user likes this object or dislikes this object. It is often impossible to determine whether the user likes this object or dislikes this object based on the state of "the user did not click and dislike this object", so the user's emotion classification for this object is often determined as neutral emotion). Among them, when the user browses short videos one by one using a short video application, if he clicks and dislikes almost every short video he browses, clicking and disliking a short video often means that he does not like the short video. If the user does not click and dislike this short video when browsing a short video later, combined with the user's habits, the user often likes this short video (because according to the user's habits, if the user does not like it, he will click and dislike it). "Not clicking and disliking this short video" can be regarded as an implicit feedback of positive emotion. However, simply based on the fact that "the user did not like the short video", the user's emotion classification for the short video is often determined as neutral, which is inconsistent with the user's positive emotion that he or she likes the short video, resulting in inaccurate emotion classification of the user for the current object.

[0210] For another example, the basic emotion classification of the manipulation behavior of "like" is positive emotion (if the user likes an object, it means that the user likes this object. If the user does not like this object, it does not mean that the user does not like this object or like this object. It is often impossible to determine whether the user likes or dislikes this object based on the state of "the user did not like this object", so the user's emotion classification for this object is often determined to be neutral emotion). Among them, when the user browses short videos one by one using a short video application, if he likes almost every short video he browses, liking a short video often means that he likes the short video. If the user does not like a short video when browsing a short video later, combined with the user's habits, the user often does not like the short video (because according to the user's habits, if the user likes it, he will like it). "Not liking the short video" can be regarded as an implicit feedback of negative emotion. However, simply based on the fact that "the user did not like the short video", the user's emotion classification for the short video is often determined as neutral, which is inconsistent with the user's positive emotion that he likes the short video, resulting in inaccurate emotion classification of the user for the current object.

[0211] Secondly, the historical manipulation behavior data of the user on the historical objects in the historical process can reflect the historical behavior habits of the user in the historical process, and the current manipulation behavior data of the user on the current object can reflect the current manipulation state of the user.

[0212] If the current manipulation state is significantly different from the user's historical behavior habits in the historical process, the user's current manipulation behavior data on the current object can often reflect the user's emotional classification of the current object as negative emotion or positive emotion.

[0213] In view of this, in the present application, when the user manipulates the current object, for each manipulation type supported by the current object, the user's current manipulation behavior data on the current object is obtained, and the user's current manipulation behavior data on the current object includes the manipulation type of the manipulation behavior actually performed by the user on the current object and the manipulation type of the manipulation behavior not performed. For each historical object that the user has manipulated in the historical process, for each manipulation type supported by the historical object, the user's historical manipulation behavior data on the historical object is obtained, and the user's historical manipulation behavior data on the historical object includes the manipulation type of the manipulation behavior actually performed by the user on the historical object and the manipulation type of the manipulation behavior not performed. Based on the user's current manipulation behavior data on the current object and the user's historical manipulation behavior data on the historical objects, the user's emotional classification of the current object is obtained.

[0214] Through this application, the user's historical manipulation behavior data on historical objects in the historical process is used as a benchmark, that is, the user's historical behavioral habits in the historical process are used as a benchmark, and the user's current manipulation behavior data on the current object are combined to obtain the user's emotional classification of the current object. Therefore, the user's emotional classification of the current object can be obtained according to the user's actual situation (such as behavioral habits, etc.), thereby improving the accuracy of the obtained user's emotional classification of the current object.

[0215] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0216] Optionally, an embodiment of the present invention further provides an electronic device, comprising: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the various processes of the above-mentioned data processing method embodiment are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0217] The embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, each process of the above-mentioned data processing method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it is not repeated here. The computer-readable storage medium is, for example, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0218] Figure 6 800 is a block diagram of an electronic device 800 shown in the present application. For example, the electronic device 800 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, etc.

[0219] Reference Figure 6 , the electronic device 800 may include one or more of the following components: a processing component 802 , a memory 804 , a power component 806 , a multimedia component 808 , an audio component 810 , an input / output (I / O) interface 812 , a sensor component 814 , and a communication component 816 .

[0220] The processing component 802 generally controls the overall operation of the electronic device 800, such as operations associated with display, phone calls, data communications, camera operations, and recording operations. The processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the above method. In addition, the processing component 802 may include one or more modules to facilitate the interaction between the processing component 802 and other components. For example, the processing component 802 may include a multimedia module to facilitate the interaction between the multimedia component 808 and the processing component 802.

[0221] The memory 804 is configured to store various types of data to support operations on the device 800. Examples of such data include instructions for any application or method operating on the electronic device 800, contact data, phone book data, messages, images, videos, etc. The memory 804 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0222] The power supply component 806 provides power to the various components of the electronic device 800. The power supply component 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the electronic device 800.

[0223] The multimedia component 808 includes a screen that provides an output interface between the electronic device 800 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 808 includes a front camera and / or a rear camera. When the device 800 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera may receive external multimedia data. Each front camera and the rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.

[0224] The audio component 810 is configured to output and / or input audio signals. For example, the audio component 810 includes a microphone (MIC), and when the electronic device 800 is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in the memory 804 or sent via the communication component 816. In some embodiments, the audio component 810 also includes a speaker for outputting audio signals.

[0225] I / O interface 812 provides an interface between processing component 802 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include but are not limited to: home button, volume button, start button, and lock button.

[0226] The sensor assembly 814 includes one or more sensors for providing various aspects of status assessment for the electronic device 800. For example, the sensor assembly 814 can detect the open / closed state of the device 800, the relative positioning of components, such as the display and keypad of the electronic device 800, and the sensor assembly 814 can also detect the position change of the electronic device 800 or a component of the electronic device 800, the presence or absence of contact between the user and the electronic device 800, the orientation or acceleration / deceleration of the electronic device 800, and the temperature change of the electronic device 800. The sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly 814 may also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 814 may also include an accelerometer, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0227] The communication component 816 is configured to facilitate wired or wireless communication between the electronic device 800 and other devices. The electronic device 800 can access a wireless network based on a communication standard, such as WiFi, a carrier network (such as 2G, 3G, 4G or 5G), or a combination thereof. In an exemplary embodiment, the communication component 816 receives a broadcast signal or broadcast operation information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 816 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

[0228] In an exemplary embodiment, the electronic device 800 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above methods.

[0229] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 804 including instructions, and the instructions can be executed by a processor 820 of an electronic device 800 to perform the above method. For example, the non-transitory 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, etc.

[0230] Figure 7 1 is a block diagram of an electronic device 1900 shown in the present application. For example, the electronic device 1900 may be provided as a server.

[0231] Reference Figure 7 , the electronic device 1900 includes a processing component 1922, which further includes one or more processors, and a memory resource represented by a memory 1932 for storing instructions executable by the processing component 1922, such as an application. The application stored in the memory 1932 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 1922 is configured to execute instructions to perform the above method.

[0232] The electronic device 1900 may also include a power supply component 1926 configured to perform power management of the electronic device 1900, a wired or wireless network interface 1950 configured to connect the electronic device 1900 to a network, and an input / output (I / O) interface 1958. The electronic device 1900 may operate based on an operating system stored in the memory 1932, such as Windows Server™, Mac OS X™, Unix™, Linux™, FreeBSD™ or the like.

[0233] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0234] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0235] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are within the protection of the present invention.

[0236] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the embodiments of the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0237] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0238] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0239] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0240] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0241] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical disks.

[0242] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A data processing method, characterized in that: Applied to a terminal, the method comprises: When a user manipulates a current object, for each manipulation type supported by the current object, obtaining current manipulation behavior data of the user on the current object, wherein the current manipulation behavior data includes the manipulation type of the manipulation behavior actually performed by the user on the current object and the manipulation type of the manipulation behavior not performed; For each historical object manipulated by the user in the historical process, for each manipulation type supported by the historical object, obtaining the historical manipulation behavior data of the user on the historical object, the historical manipulation behavior data including the manipulation type of the manipulation behavior actually performed by the user on the historical object and the manipulation type of the manipulation behavior not performed; According to the current manipulation behavior data and the historical manipulation behavior data, the emotion classification of the user towards the current object is obtained, and the execution result of whether the user has performed the manipulation behavior of the manipulation type on the current object is determined through the current manipulation behavior data. The statistical result of whether the user has an execution tendency to perform the manipulation behavior of the manipulation type on the historical object is obtained through the historical manipulation behavior data. There is a basic emotion classification for the manipulation behavior of the manipulation type. The emotion classification is determined based on the basic emotion classification of the manipulation behavior of the manipulation type, the statistical results and the execution results, and the emotion score of the user for the manipulation behavior of the current object regarding the manipulation type is obtained.

2. The method according to claim 1, characterized in that The acquiring, according to the current manipulation behavior data and the historical manipulation behavior data, the emotion classification of the user toward the current object comprises: Acquire a sentiment classification acquisition model, wherein the sentiment classification acquisition model is obtained by training based on a plurality of training data sets, wherein the training data sets include: first sample manipulation behavior data of a sample user on a first sample object, second sample manipulation behavior data of the sample user on a second sample object, and annotated sentiment classification of the sample user on the second sample object; Inputting the current manipulation behavior data and the historical manipulation behavior data into the emotion classification acquisition model, so that the emotion classification acquisition model processes the current manipulation behavior data and the historical manipulation behavior data, obtains the user's emotion classification of the current object, and outputs the user's emotion classification of the current object; Acquire the sentiment classification of the user to the current object output by the sentiment classification acquisition model.

3. The method according to claim 1, characterized in that The acquiring of the sentiment classification acquisition model comprises: According to the historical manipulation behavior data, an emotion classification acquisition model suitable for the user is acquired from a plurality of trained emotion classification acquisition models.

4. The method according to claim 3, characterized in that: The step of acquiring a sentiment classification acquisition model suitable for the user from a plurality of trained sentiment classification acquisition models according to the historical manipulation behavior data includes: Determining, from a plurality of user sets, a user set to which the user belongs according to the historical manipulation behavior data; Among the sentiment classification acquisition models respectively applicable to each user set, a sentiment classification acquisition model applicable to the user set to which the user belongs is selected.

5. The method according to claim 1, characterized in that The acquiring, according to the current manipulation behavior data and the historical manipulation behavior data, the emotion classification of the user toward the current object comprises: For any manipulation behavior of a manipulation type supported by the current object, determine the current execution status of the manipulation behavior of the current object with respect to the manipulation type by the user according to the current manipulation behavior data, and determine the historical execution status of the manipulation behavior of the historical object with respect to the manipulation type by the user according to the historical manipulation behavior data, and obtain the emotion score of the user with respect to the manipulation behavior of the current object with respect to the manipulation type according to the current execution status and the historical execution status; The emotion classification of the user on the current object is obtained according to the emotion scores of the user's manipulation behaviors on the current object regarding each manipulation type.

6. The method according to claim 1, characterized in that The statistical result of counting, based on the historical execution situation, whether the user has an execution tendency to execute the manipulation behavior of the manipulation type on the historical object includes: Acquire the historical number of historical objects manipulated by the user in the historical process, and acquire the number of times the user performs manipulation actions of the manipulation type on the historical objects in the process of manipulating the historical objects in the historical process; A statistical result of whether the user has an execution tendency to execute the manipulation behavior of the manipulation type on the historical object is counted according to the history quantity and the execution times.

7. The method according to claim 6, characterized in that The statistical result of counting, based on the history quantity and the execution times, whether the user has an execution tendency to execute the manipulation behavior of the manipulation type on the history object includes: Calculating a ratio between the number of executions and the number of history; When the ratio is greater than or equal to a preset ratio, determining that the user has an execution tendency to execute a manipulation behavior of the manipulation type on the historical object; or, When the ratio is smaller than a preset ratio, it is determined that the user does not have an execution tendency to execute a manipulation behavior of the manipulation type on the historical object.

8. The method according to claim 1, characterized in that The obtaining, according to the basic emotion classification of the manipulation behavior of the manipulation type, the statistical result and the execution result, the emotion score of the user for the manipulation behavior of the current object with respect to the manipulation type includes: In the correspondence between the basic emotion classification of the manipulation behavior, the statistical results of the execution tendency of the manipulation behavior, the execution results of the manipulation behavior and the user's emotion score for the manipulation behavior, find the user's emotion score for the manipulation behavior that corresponds to the basic emotion classification of the manipulation behavior of the manipulation type, the statistical results and the execution results, and use it as the user's emotion score for the manipulation behavior of the current object regarding the manipulation type.

9. The method according to claim 5, characterized in that The acquiring the user's emotion classification of the current object according to the emotion scores of the user's manipulation behaviors of each manipulation type on the current object includes: Taking a weighted sum of the emotion scores of the user's manipulation behaviors on the current object for each manipulation type, to obtain a total emotion score of the user on the current object; The emotion classification of the user towards the current object is obtained according to the total emotion value of the user towards the current object.

10. The method according to claim 9, characterized in that The obtaining the emotion classification of the user towards the current object according to the total emotion value of the user towards the current object comprises: In a case where the total emotion value of the user towards the current object is greater than a preset emotion threshold, determining that the emotion of the user towards the current object is classified as a positive emotion; or, When the total emotion value of the user towards the current object is equal to a preset emotion threshold, determining that the emotion of the user towards the current object is classified as a neutral emotion; or, When the total value of the user's emotion toward the current object is less than a preset emotion threshold, it is determined that the user's emotion toward the current object is classified as a negative emotion.

11. A data processing device, characterized in that: Applied to a terminal, the device comprises: A first acquisition module is used to acquire, when a user manipulates a current object, current manipulation behavior data of the user on the current object for each manipulation type supported by the current object, wherein the current manipulation behavior data includes the manipulation type of the manipulation behavior actually performed by the user on the current object and the manipulation type of the manipulation behavior not performed; A second acquisition module is used to acquire, for each historical object that the user has manipulated in the historical process, historical manipulation behavior data of the user on the historical object for each manipulation type supported by the historical object, wherein the historical manipulation behavior data includes the manipulation type of the manipulation behavior actually performed by the user on the historical object and the manipulation type of the manipulation behavior not performed; The third acquisition module is used to obtain the user's emotion classification of the current object based on the current manipulation behavior data and the historical manipulation behavior data, determine whether the user has performed the execution result of the manipulation behavior of the manipulation type on the current object through the current manipulation behavior data, and obtain statistical results of whether the user has the execution tendency to perform the manipulation behavior of the manipulation type on the historical object through the historical manipulation behavior data. The manipulation behavior of the manipulation type has a basic emotion classification, and the emotion classification is determined based on the basic emotion classification of the manipulation behavior of the manipulation type, the statistical results and the execution results, and the emotion score of the user's emotion towards the current object regarding the manipulation behavior of the manipulation type.

12. An electronic device, characterized in that: include: A processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the steps of the data processing method according to any one of claims 1 to 10 when executed by the processor.

13. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the data processing method according to any one of claims 1 to 10 are implemented.

Images