Method of classifying data, related apparatus and computer program product

Abstract

The invention provides a data classification method, a related device and a computer program product, according to the application, based on data features of to-be-classified data, a preliminary classification result corresponding to the to-be-classified data and classification confidence for the preliminary classification result are generated, and the preliminary classification result comprises a first data type or a second data type; in response to the fact that the classification confidence coefficient is larger than or equal to a first confidence coefficient threshold value and smaller than or equal to a second confidence coefficient threshold value, obtaining associated information of an uploading user of the to-be-classified data; in response to the situation that the uploading user is determined as a target user based on the associated information, the to-be-classified data is classified into a first data type, and the number of historical data of the first data type released by the target user in the historical time period is larger than or equal to a number threshold value. Therefore, while the data classification efficiency is considered, the data classification quality is improved, and the method can be applied to more classification scenes.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method and device for classifying data, an electronic device, a computer-readable medium, and a computer program product. Background Art

[0002] With the development of society, Internet technology is also improving day by day. As a platform that provides users with resources and opportunities to communicate with other users, the Internet allows users to obtain the resources they need through the Internet, and communicate, exchange and share content with other users.

[0003] For example, users can obtain the required data from the platform or other users through the Internet for use. For example, the data can be video data, and users can obtain these video data through the platform to obtain content. Correspondingly, when providing these data to other users, in order to improve the efficiency of data management and sharing, it is often necessary to classify these data so as to adaptively configure management and sharing strategies based on the classification results, thereby improving the efficiency and strategies of management and sharing. Therefore, how to classify videos more efficiently and with higher quality is worthy of attention and urgent need. Summary of the invention

[0004] Multiple aspects of the present application provide a method, device, electronic device, computer-readable storage medium and computer program product for classifying data, which can improve the classification quality of data while taking into account the efficiency of data classification.

[0005] In one aspect of the present application, a method for classifying data is provided, comprising: generating a preliminary classification result corresponding to the data to be classified and a classification confidence for the preliminary classification result based on data features of the data to be classified, wherein the preliminary classification result includes a first data type or a second data type; in response to the classification confidence being greater than or equal to a first confidence threshold and less than or equal to a second confidence threshold, obtaining associated information of a user who uploaded the data to be classified; in response to determining that the uploading user is a target user based on the associated information, classifying the data to be classified into a first data type, wherein the number of historical data of the first data type published by the target user within a historical period is greater than or equal to a quantity threshold.

[0006] In another aspect of the present application, there is provided an apparatus for classifying data, comprising: a data feature processing module, configured to generate a preliminary classification result corresponding to the data to be classified and a classification confidence for the preliminary classification result based on the data features of the data to be classified, wherein the preliminary classification result includes a first data type or a second data type; an associated information acquisition module, configured to acquire associated information of an uploading user of the data to be classified in response to the classification confidence being greater than or equal to a first confidence threshold and less than or equal to a second confidence threshold; a first data classification module, configured to classify the data to be classified into a first data type in response to determining that the uploading user is a target user based on the associated information, wherein the number of historical data of the first data type published by the target user within a historical period is greater than or equal to a quantity threshold.

[0007] Another aspect of the present application provides an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method for classifying data provided above.

[0008] In another aspect of the present application, a computer-readable storage medium is provided, on which computer program instructions are stored. The computer program instructions can be executed by a processor to implement the method for classifying data provided above.

[0009] In another aspect of the present application, a computer program product includes a computer program having computer program instructions stored thereon. When the computer program is executed by a processor, the method for classifying data as provided above can be implemented.

[0010] In the solution provided by the embodiment of the present application, based on the data features of the data to be classified, a preliminary classification result corresponding to the data to be classified and a classification confidence for the preliminary classification result are generated, wherein the preliminary classification result includes the first data type or the second data type; in response to the classification confidence being greater than or equal to the first confidence threshold and less than or equal to the second confidence threshold, the associated information of the uploading user of the data to be classified is obtained; in response to determining that the uploading user is the target user based on the associated information, the data to be classified is classified into the first data type, wherein the number of historical data of the first data type released by the target user in the historical period is greater than or equal to the number threshold. In this way, the data classification efficiency can be taken into account while improving the classification quality of the data, and it can be applied to more classification scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present application, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0012] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings:

[0013] Figure 1 A flowchart of a data classification process provided in an embodiment of the present application;

[0014] Figure 2 A flowchart of a process of classifying data implemented in a specific application scenario provided by an embodiment of the present application;

[0015] Figure 3 A schematic diagram of the structure of a device for classifying data provided in one embodiment of the present application;

[0016] Figure 4 The present invention is a schematic diagram of the structure of an electronic device suitable for implementing the solution in the embodiment of the present application.

[0017] The same or similar reference numerals in the drawings represent the same or similar components. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0019] In a typical configuration of the present application, the terminal and the equipment of the service network each include one or more processors (CPU), input / output interface, network interface and memory.

[0020] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0021] Computer readable media include permanent and non-permanent, removable and non-removable media, and can be implemented by any method or technology to store information. Information can be computer program instructions, data structures, modules of programs or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.

[0022] As discussed above, how to classify videos more efficiently and with higher quality is worthy of attention and urgent need.

[0023] In some schemes, the data to be classified will be processed to obtain data features. Then, the data type corresponding to the data to be classified will be determined based on the similarity matching results between the "standard" and "reference" data features corresponding to each data type and the data features.

[0024] In such an approach, a "similarity" threshold is often chosen to determine whether a "standard" or "reference" data feature matches the data feature (for example, determining whether the similarity threshold is greater than a predetermined similarity threshold).

[0025] In such cases, in order to ensure the quality of classification, only those data to be classified that provide very high quality "data features" can be processed. As a result, this method has limited data classification capabilities, not only making it difficult to balance data classification quality and classification efficiency, but also limiting actual specific scenarios.

[0026] In this regard, an embodiment of the present application provides a method for classifying data, which generates a preliminary classification result corresponding to the data to be classified and a classification confidence for the preliminary classification result based on the data features of the data to be classified, wherein the preliminary classification result includes a first data type or a second data type; in response to the classification confidence being greater than or equal to a first confidence threshold and less than or equal to a second confidence threshold, obtain the associated information of the uploading user of the data to be classified; in response to determining that the uploading user is a target user based on the associated information, classify the data to be classified into the first data type, wherein the number of historical data of the first data type published by the target user in the historical period is greater than or equal to the quantity threshold. In this way, the data classification efficiency can be taken into account while improving the classification quality of the data, and it can be applied to more classification scenarios.

[0027] In actual scenarios, the execution subject of the method can be a user device, or a device formed by integrating a user device and a network device through a network, or an application running on the above device. The user device includes but is not limited to various terminal devices such as computers, mobile phones, tablet computers, smart watches, and bracelets. The network device includes but is not limited to network hosts, single network servers, multiple network server sets, or cloud computing-based computer sets. Here, the cloud is composed of a large number of hosts or network servers based on cloud computing (Cloud Computing), where cloud computing is a type of distributed computing, a virtual computer composed of a group of loosely coupled computer sets.

[0028] When the execution subject is software, it can be installed in the electronic devices listed above. It can be implemented as multiple software or software modules, or as a single software or software module, and no specific limitation is made here.

[0029] For ease of understanding, a "server" that can provide "data" to users online will be used as an example of an execution subject. For example, the data to be classified can be collected by the server and can be subsequently provided to the terminal device used by the user. For example, the data to be classified can be a video uploaded by the user to the server. Accordingly, the server can classify the video, adopt different publishing and sharing strategies based on the classification results, and then actually take action according to the publishing and sharing strategies.

[0030] Figure 1 A process 100 for classifying data provided by an embodiment of the present application is shown. The process 100 includes at least the following processing steps:

[0031] (Step) S101, based on the data features of the data to be classified, generate a preliminary classification result corresponding to the data to be classified and a classification confidence level for the preliminary classification result.

[0032] In an embodiment of the present application, the execution entity, after acquiring the data to be classified, can generate data features of the data to be classified, and based on the data features of the data to be classified, generate a preliminary classification result corresponding to the data to be classified and a classification confidence level for the preliminary classification result.

[0033] The data to be classified may be, for example, data in the modes of images, text, audio, etc. Correspondingly, for unimodal data to be classified, the execution subject may directly process it to obtain corresponding data features. For example, when the data to be classified is an image, the execution subject may choose to generate the "data features" of the image. For example, based on neural networks such as VGG (Visual Geometry Group), MobileNet, ViT (Vision Transformer), etc., the image is processed to obtain corresponding data features. For example, data features may correspondingly describe the "content" of the data to be classified in this mode. For example, "content" may be characters, scenes, logos, emotional polarity, action intentions, etc. included in the data to be classified.

[0034] Then, the execution subject can obtain the preliminary classification results corresponding to the data to be classified and the classification confidence for the preliminary classification results based on the data features, or in other words, process these data features.

[0035] The preliminary classification result may be one of the first data type or the second data type, and the first data type and the second data type may be determined based on the pre-classification requirements. For example, the first data type may be content that is not allowed to be published or shared (for example, data under this type may be data that other users do not want to see), while the second data type may be content that is allowed to be published or shared. For ease of understanding, the following will also be explained for exemplary purposes using such data classification criteria.

[0036] For example, for an image, the execution entity may determine, based on data features, or image features, whether the image includes image content that is not desired or is not allowed to be published or shared.

[0037] If such content exists, the execution subject may choose to preliminarily classify it into the first data type, that is, determine a preliminary classification result (or, may be referred to as a preliminary classification result). Moreover, after determining the preliminary classification result, the execution subject may also output a "confidence" for the preliminary classification result, that is, the execution subject's degree of trust in the output preliminary classification result.

[0038] For example, in the process of processing data features using a neural network and a model, a "classification head" for generating confidence can be added to the neural network to supervise the processing process and generate corresponding (classification) confidence. For example, the configured classification head can generate confidence by using a loss function. The loss function can convert the original prediction score of the neural network into a probability distribution to represent the probability of each data type.

[0039] Accordingly, the classification head may choose to use these probability values ​​as confidence levels to “characterize” the “reliability” of the model in classifying the data to be classified into the corresponding data type.

[0040] In this way, the type corresponding to the value with the highest final probability can be used as the "preliminary classification result", and the corresponding probability value is correspondingly determined as the classification confidence.

[0041] In some embodiments, the execution entity may choose to utilize the complete model to continuously implement the actions of processing the data to be classified, generating data features of the data to be classified, and processing the data features to generate preliminary classification results corresponding to the data to be classified and classification confidence levels for the preliminary classification results.

[0042] As a result, not only can the configuration difficulty of the execution entity be simplified, but also through the configuration and training of these models, the models can be used to fully learn how to generate data features, what kind of data features to generate, and how to use these data features, so that the models can be used to more accurately determine the preliminary classification results corresponding to the data to be classified and the classification confidence of the preliminary classification results.

[0043] Specifically, in the process of generating the preliminary classification result corresponding to the data to be classified and the classification confidence for the preliminary classification result based on the data features of the data to be classified, the execution subject may choose to use the data classification model to process the data to be classified to generate the data features of the data to be classified. Then, after the data features are generated, the data classification model is used to process the data features to generate the preliminary classification result corresponding to the data to be classified and the classification confidence for the preliminary classification result.

[0044] S102: In response to the classification confidence being greater than or equal to a first confidence threshold and less than or equal to a second confidence threshold, obtaining association information of a user who uploaded the data to be classified.

[0045] In an embodiment of the present application, after the classification confidence is obtained based on the above S101, it is compared with the first confidence threshold and the second confidence. The first confidence threshold can be determined based on the standard of "credibly unreliable" of the preliminary classification result. That is, if the classification confidence is lower than the first confidence threshold, it can be determined that the preliminary classification result is credibly unreliable.

[0046] Similarly, for the second confidence threshold, it can be determined based on the criterion of the preliminary classification result being “credibly reliable.” That is, if the classification confidence is higher than the second confidence threshold, it can be determined that the preliminary classification result is credibly reliable.

[0047] If the classification confidence is greater than or equal to the first confidence threshold and less than or equal to the second confidence threshold, that is, based on the classification confidence, the preliminary classification result can neither be credibly considered to be reliable nor be credibly considered to be unreliable, then in such a situation, the execution entity can respond to this and choose to obtain the associated information of the uploading user to determine and identify whether the uploading user of the data to be classified is the target user.

[0048] That is, when it is impossible to reliably determine what type of data the data to be classified should be classified into based solely on data features, the executing entity may choose to introduce the associated information of the uploading user who uploaded the data to be classified for further judgment and determination, or to assist in determining the classification tendency of the "data to be classified".

[0049] Accordingly, the execution subject can use the association information to determine whether the uploading user is the target user. If the number of historical data of the first data type published by the uploading user during the historical period is greater than or equal to the number threshold, the execution subject can determine that the uploading user is the target user.

[0050] That is, if the uploading user publishes historical data classified as the first data type at least a number of times corresponding to the quantity threshold within the historical period, the uploading user may be determined to be a “target user”.

[0051] For example, for user A, among the historical data provided by the user in the historical period, there may be B pieces classified as the first data type. If the “B” is greater than or equal to the quantity threshold C, the user A may be determined to be the target user.

[0052] In some embodiments, the associated information includes at least the number of historical data of the first data type published by the uploading user within the historical time period, and the execution entity can determine whether the uploading user is the target user based on a comparison between the number and a number threshold.

[0053] In some embodiments, the associated information may also include indicators such as the registration time of the uploading user's account, the activity level in the historical period (for example, the number of logins, the number of interactions with other users), and the number of uploads of historical data. Accordingly, the execution entity may further "dynamically" determine, for example, the "historical period" and "quantity threshold" based on these indicators. For example, the execution entity may determine the selection range of the above-mentioned "historical period" (for example, the length, specific time interval) based on the registration time, or may determine the corresponding quantity threshold based on the activity level. As a result, the "historical period", "quantity threshold", etc. may be "dynamically" adjusted according to the different situations of different users. As a result, it is possible to more accurately determine whether the uploading user is a "target user".

[0054] In some embodiments, similarly, the "indicator" can also be used by adjusting the quantity threshold corresponding to the quantity of historical data. For example, for an uploader E, if his registration time is earlier than that of another uploader F, the quantity threshold corresponding to the uploader E may be smaller than the quantity threshold corresponding to the uploader F. Thus, the above standards can be adjusted flexibly and dynamically by referring to the historical behavior of the uploader.

[0055] S103: In response to determining that the uploading user is a target user based on the association information, classify the data to be classified into a first data type.

[0056] In an embodiment of the present application, if the execution subject determines that the uploading user is the target user, the execution subject may classify the data to be classified into the first data type based on a “reference” to this.

[0057] Therefore, when the classification results obtained from processing the features of the data to be classified are not "credibly reliable" or "credibly unreliable", the executing entity may choose to refer to the user's historical behavior and manage the data to be classified uploaded by the user according to the first data type to avoid its unexpected outflow.

[0058] Then, the method for classifying data provided by the present application generates a preliminary classification result corresponding to the data to be classified and a classification confidence for the preliminary classification result based on the data features of the data to be classified, wherein the preliminary classification result includes the first data type or the second data type; in response to the classification confidence being greater than or equal to the first confidence threshold and less than or equal to the second confidence threshold, obtain the associated information of the uploading user of the data to be classified; in response to determining that the uploading user is the target user based on the associated information, classify the data to be classified into the first data type, wherein the number of historical data of the first data type published by the target user in the historical period is greater than or equal to the quantity threshold. In this way, the data classification efficiency can be taken into account while improving the classification quality of the data, and it can be applied to more classification scenarios.

[0059] In some embodiments, in response to the classification confidence being greater than a second threshold, the preliminary classification result is determined as the data type of the data to be classified.

[0060] Specifically, if the execution subject determines that the classification confidence is greater than the second threshold, that is, the preliminary classification result obtained based only on the data features is "credibly reliable", then the execution subject can choose to determine the preliminary classification result as the data type of the data to be classified. Therefore, for those data to be classified whose "data type" can be "credibly and reliably" determined locally by the execution subject, the classification is completed directly to improve the classification efficiency. (Because in this case, no additional reference is made to "more information" such as associated information).

[0061] In some embodiments, if the classification confidence is less than the first threshold, the execution subject may choose to provide the data to be classified to the target device based on a preset communication path, so as to classify the data to be classified using the target device. For example, the target device may be another server that is deployed with a model with stronger performance and stronger performance for data feature processing, or the target device may also be a terminal device used by a classifier, etc. Thus, the classification results are provided by using these servers and terminal devices with "stronger" capabilities.

[0062] Accordingly, if the target device returns an indication of the classification result for the data to be classified, the execution subject can choose to determine the indication of the classification result as the data type of the data to be classified. Thus, for the classification result that is not reliable for the execution subject, the execution subject can realize the classification by collaboratively calling other target devices, so as to further improve the data classification capability and ensure the classification quality.

[0063] Based on any of the above embodiments, as described above, the data uploaded by the user may be multimodal. That is, in order to enrich the user's content sharing and acquisition experience, the user may expect to upload and obtain "multimodal" data. For example, the data may actually be "multimodal data" composed of at least two of images, text, and audio.

[0064] Accordingly, if the original data acquired by the execution subject includes data to be classified in at least two modes, the execution subject may choose to split the data in each mode from the original data based on the mode to obtain the data to be classified in each mode. That is, each data to be classified is data in a separate mode.

[0065] Then, the execution subject may similarly perform the processing steps discussed above for the data to be classified in each modality to determine the data type corresponding to each data to be classified.

[0066] In some embodiments, for data to be classified in different modalities, the execution entity may choose to call the classification model of the corresponding modality to process the data to be classified in the corresponding modality to obtain a preliminary classification result corresponding to the data to be classified and a classification confidence level for the preliminary classification result.

[0067] Then, the execution subject may refer to the process discussed above and similarly determine the final classification results of the data to be classified under each modality.

[0068] Next, for the final classification result, if among the data types corresponding to the data to be classified under each modality, there is data to be classified of the first data type (or in other words, among the data to be classified under each modality, there is at least one data to be classified under the first data type that is ultimately classified as the first data type), the execution entity may choose to determine the data type of the entire original data as the first data type.

[0069] This not only enables classification and management of multimodal data, but also avoids limiting multimodal data that includes potentially undesirable content in at least one modality.

[0070] In some embodiments, if the execution subject determines that there is no data of the first data type in the data types corresponding to the data to be classified in each modality (or in other words, the data to be classified in each modality are finally classified as the second data type), the execution subject may choose to classify the data type of the original data as the second data type, thereby allowing the publication and sharing of the multimodal data.

[0071] In some embodiments, if the execution subject uses the classification model to process the classification data to generate data features, and uses the data classification model to process the data features to generate the preliminary classification results corresponding to the data to be classified and the classification confidence for the preliminary classification results, in such a case, the execution subject may also have the "update capability" for the classification model. Thus, the execution subject can "iterate" the classification model's capabilities based on the update of the classification model, so that the classification model has better processing performance and can complete the classification with higher quality.

[0072] Specifically, the execution entity can also provide the data to be classified to the target device (for example, the server deployed with a more powerful model discussed above, the terminal device used by the classifier, etc.) based on a preset communication path.

[0073] Accordingly, if the target device returns updated data features for the data to be classified, the execution subject can respond to this, use the data to be classified as input, and the updated data features as output, to update and train the data classification model. In this way, the execution subject can use the "updated data features" provided by the target device to update and iterate the classification model, thereby improving the performance of the device.

[0074] In some embodiments, the data to be classified may correspond to more than two data features at the same time. That is, in the process of generating the preliminary classification result corresponding to the data to be classified and the classification confidence for the preliminary classification result based on the data features of the data to be classified, the execution subject may actually use at least two data features of the data to be classified.

[0075] In such a case, "updating data features" may only be for part of the multiple data features. For example, in the case of corresponding to more than two data features at the same time, "updating data features" may only be for one of the data features. Therefore, the execution subject can also use the "updating data features" to update only the processing capability of the classification model corresponding to the "updated data features".

[0076] Similarly, in some embodiments, if the data to be classified corresponds to more than two data features at the same time, the execution subject can also select some of the data features for actual use through a screening strategy. For example, the execution subject can determine the use value of each data feature based on the scenario corresponding to the data to be classified, and only use the data features whose use value exceeds the value threshold.

[0077] In some embodiments, if the uploading user is not the target user, the execution entity may choose to use the preliminary classification result as the "temporary result" of the data to be classified. In other words, the execution entity may choose to classify it as the result indicated by the preliminary classification result, but may publish or provide it to the classifier by adding "prompt information" to it, thereby prompting other users and classifiers.

[0078] Next, the execution subject can adjust the long-term classification of the data to be classified based on the feedback and instructions from other users and classifiers. For example, for temporarily classifying the data to be classified as the second data type, the execution subject can choose to publish it to some other users (for example, expert users who can provide classification suggestions) in a short term or in a targeted manner. Accordingly, the execution subject can determine the long-term data type of the data to be classified based on the feedback from other users in the "short term".

[0079] It should be understood that in the technical solutions involved in this application, the acquisition, storage, use, processing, transportation, provision and disclosure of the user personal information involved (such as the indexes, data, user-related information, etc. provided and fed back by the users involved in the above application) are in compliance with the relevant laws and regulations and do not violate public order and good morals.

[0080] For easier understanding, you can also refer to Figure 2 . Figure 2 The flowchart of the process 200 of classifying data provided by the embodiment of the present application and implemented in a specific application scenario is shown. The process 200 is specifically as follows:

[0081] First, an uploading user (not shown) can execute S201 by using a terminal device 201 to communicate with a server 220 and upload original data 215 to the server 220. Exemplarily, the original data 215 is "multimodal data" including data in three modalities: image, text, and audio.

[0082] After receiving the original data 215, the server 220 executes S202 to split the original data 215 into three modes: image, text, and audio, to obtain single-mode data to be classified, for example, data to be classified 221 in image mode, data to be classified 222 in text mode, and data to be classified 223 in audio mode.

[0083] Next, the server 220 calls the classification model corresponding to each modality respectively to determine the preliminary classification result and classification confidence corresponding to each data to be classified. As shown in the figure, the server 220 can call the classification model 231, process the data to be classified 221, and obtain the corresponding initial classification result 241 (exemplarily, it can indicate that the data to be classified 221 is the second data type) and the corresponding classification confidence 251; the server 220 can call the classification model 232, process the data to be classified 222, and obtain the corresponding initial classification result 242 (exemplarily, it can indicate that the data to be classified 222 is the second data type) and the corresponding classification confidence 252; the server 220 can call the classification model 233, process the data to be classified 223, and obtain the corresponding initial classification result 243 (exemplarily, it can indicate that the data to be classified 223 is the second data type) and the corresponding classification confidence 253.

[0084] Exemplarily, the classification confidences 251 and 253 are both greater than the second confidence threshold, and accordingly, the server 220 may classify the to-be-classified data 221 and 223 as the second data type based on steps S203 - 1 and S203 - 2 .

[0085] Exemplarily, the classification confidence 252 is greater than or equal to the first confidence threshold, and less than or equal to the second confidence threshold. In response, the server 220 executes S204 to obtain the associated information of the uploading user (for example, obtain the number of historical data of the first data type published by the uploading user in the historical period, as recorded in the upload record of the terminal device 110). Exemplarily, the number of historical data of the first data type published by the (uploading) user using the terminal device 110 in the historical period is greater than or equal to the quantity threshold.

[0086] Accordingly, the server 220 may execute S205 to determine that the uploading user (using the uploading of the terminal device 110 ) is the target user.

[0087] In this regard, the server 220 may continue to execute S206 to classify the to-be-classified data 222 into the first data type.

[0088] Next, because the data to be classified 222 in the original data 215 is classified as the first data type, the server 220 may execute S207 to finally classify the original data 215 as the first data type.

[0089] The present application also provides a device for classifying data. The structure of the device is as follows: Figure 3 The device 300 shown. The device 300 includes: a data feature processing module 310, configured to generate a preliminary classification result corresponding to the data to be classified and a classification confidence for the preliminary classification result based on the data features of the data to be classified, wherein the preliminary classification result includes a first data type or a second data type; an associated information acquisition module 320, configured to obtain the associated information of the uploading user of the data to be classified in response to the classification confidence being greater than or equal to the first confidence threshold and less than or equal to the second confidence threshold; a first data classification module 330, configured to classify the data to be classified into a first data type in response to determining that the uploading user is a target user based on the associated information, wherein the number of historical data of the first data type published by the target user in the historical period is greater than or equal to the quantity threshold.

[0090] In some embodiments, the apparatus 300 further includes: a second data classification module configured to determine the preliminary classification result as the data type of the data to be classified in response to the classification confidence being greater than a second threshold.

[0091] In some embodiments, the device 300 also includes: a first data providing module, configured to provide the data to be classified to the target device based on a preset communication path in response to the classification confidence being less than a first threshold; and a third data classification module, configured to determine the indicated classification result as the data type of the data to be classified in response to the target device returning an indicated classification result for the data to be classified.

[0092] In some embodiments, the device 300 also includes: an original data acquisition module, configured to acquire original data, wherein the original data includes data to be classified of at least two modalities; a data extraction module to be classified, configured to split the data to be classified in each modality from the original data; and a first original data type determination module, configured to classify the data type of the original data into the first data type in response to the presence of data to be classified of the first data type.

[0093] In some embodiments, the apparatus 300 further includes: a second original data type determination module configured to classify the data type of the original data into a second data type in response to the absence of to-be-classified data of the first data type.

[0094] In some embodiments, the data feature processing module 310 includes: a feature generation submodule, configured to use a data classification model to process the data to be classified and generate data features of the data to be classified; a feature processing submodule, configured to use a data classification model to process data features and generate preliminary classification results corresponding to the data to be classified and classification confidence for the preliminary classification results.

[0095] In some embodiments, the apparatus 300 further includes: a second data providing module, configured to provide the data to be classified to the target device based on a preset communication path; and a model updating training module, configured to update and train the data classification model in response to the target device returning updated data features for the data to be classified, using the data to be classified as input and the updated data features as output.

[0096] Based on the same inventive concept, an electronic device, a readable storage medium and a computer program product are also provided in an embodiment of the present application. The method corresponding to the electronic device may be the method for classifying data in the aforementioned embodiment, and its principle of solving the problem is similar to that of the method. The electronic device provided in an embodiment of the present application includes: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions executable by at least one processor, and the instructions are executed by at least one processor so that at least one processor can execute the methods and / or technical solutions of the aforementioned multiple embodiments of the present application.

[0097] The electronic device may be a user device, or a device formed by integrating a user device and a network device through a network, or an application running on the above device. The user device includes but is not limited to various terminal devices such as computers, mobile phones, tablet computers, smart watches, and bracelets. The network device includes but is not limited to network hosts, single network servers, multiple network server sets, or cloud computing-based computer sets, which can be used to implement some processing functions when setting an alarm. Here, the cloud is composed of a large number of hosts or network servers based on cloud computing, where cloud computing is a type of distributed computing, a virtual computer composed of a group of loosely coupled computer sets.

[0098] Figure 4The structure of an electronic device suitable for implementing the method and / or technical solution in the embodiment of the present application is shown, and the electronic device 400 includes a central processing unit (CPU, Central Processing Unit) 401, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM, Read Only Memory) 402 or the program loaded from the storage part 408 to the random access memory (RAM, Random Access Memory) 403. In RAM 403, various programs and data required for system operation are also stored. CPU 401, ROM 402 and RAM 403 are connected to each other through bus 404. Input / output (I / O, Input / Output) interface 405 is also connected to bus 404.

[0099] The following components are connected to the I / O interface 405: an input section 406 including a keyboard, a mouse, a touch screen, a microphone, an infrared sensor, etc.; an output section 407 including a cathode ray tube (CRT), a liquid crystal display (LCD), an LED display, an OLED display, etc., and a speaker, etc.; a storage section 408 including one or more computer-readable media such as a hard disk, an optical disk, a magnetic disk, a semiconductor memory, etc.; and a communication section 409 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 409 performs communication processing via a network such as the Internet.

[0100] In particular, the methods and / or embodiments in the embodiments of the present application may be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes program code for executing the method shown in the flowchart. When the computer program is executed by the central processing unit (CPU) 401, the above functions defined in the method of the present application are executed.

[0101] Another embodiment of the present application further provides a computer-readable storage medium and a computer program product, on which computer program instructions are stored. The computer program instructions can be executed by a processor to implement the methods and / or technical solutions of any one or more embodiments of the present application.

[0102] Specifically, the present embodiment can adopt any combination of one or more computer-readable media. Computer-readable media can be computer-readable signal media or computer-readable storage media. Computer-readable storage media can be, for example, systems, devices or devices including but not limited to electricity, magnetism, light, electromagnetic, infrared, or semiconductors, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: electrical connections with one or more wires, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this document, computer-readable storage media can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.

[0103] Computer readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than a computer readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0104] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0105] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0106] The flow chart or block diagram in the accompanying drawings shows the possible architecture, function and operation of the equipment, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated system for hardware that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0107] 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.

[0108] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of modules and units is only a logical function division. There may be other division methods in actual implementation. For example, with units as an example, multiple units or page 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 an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0109] 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.

[0110] In addition, each functional module or unit in each embodiment of the present application may be integrated into one processing module or unit, or each module or unit may exist physically separately, or two or more units may be integrated into one module or unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules or units.

[0111] The above-mentioned integrated modules and units implemented in the form of software function modules and units can be stored in a computer-readable storage medium. The above-mentioned software function modules and units are stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor (processor) to perform some steps of the methods of each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store program codes.

[0112] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

[0113] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in a device claim can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

Claims

1. A method for classifying data, characterized in that include: Based on data features of the data to be classified, generating a preliminary classification result corresponding to the data to be classified and a classification confidence for the preliminary classification result, wherein the preliminary classification result includes a first data type or a second data type; In response to the classification confidence being greater than or equal to a first confidence threshold and less than or equal to a second confidence threshold, acquiring association information of a user who uploaded the data to be classified; In response to determining that the uploading user is a target user based on the association information, the data to be classified is classified into the first data type, wherein the amount of historical data of the first data type published by the target user within a historical period is greater than or equal to a quantity threshold.

2. The method according to claim 1, characterized in that Also includes: In response to the classification confidence being greater than the second threshold, the preliminary classification result is determined as the data type of the data to be classified.

3. The method according to claim 1, characterized in that Also includes: In response to the classification confidence being less than the first threshold, providing the data to be classified to a target device based on a preset communication path; In response to the target device returning an indication classification result for the data to be classified, determining the indication classification result as the data type of the data to be classified.

4. The method according to any one of claims 1 to 3, characterized in that Also includes: Acquire original data, wherein the original data includes data to be classified in at least two modes; Splitting the original data into data to be classified in each mode; In response to the existence of to-be-classified data of the first data type, the data type of the original data is classified as the first data type.

5. The method according to claim 4, characterized in that Also includes: In response to the absence of to-be-classified data of the first data type, the data type of the original data is classified as the second data type.

6. The method according to claim 1, characterized in that The step of generating a preliminary classification result corresponding to the data to be classified and a classification confidence level for the preliminary classification result based on the data features of the data to be classified includes: Processing the data to be classified using the data classification model to generate data features of the data to be classified; The data features are processed using the data classification model to generate a preliminary classification result corresponding to the data to be classified and a classification confidence level for the preliminary classification result.

7. The method according to claim 6, characterized in that Also includes: Based on a preset communication path, providing the data to be classified to a target device; In response to the target device returning updated data features for the data to be classified, the data classification model is updated and trained using the data to be classified as input and the updated data features as output.

8. A device for classifying data, characterized in that include: a data feature processing module configured to generate a preliminary classification result corresponding to the data to be classified and a classification confidence for the preliminary classification result based on the data feature of the data to be classified, wherein the preliminary classification result includes the first data type or the second data type; an associated information acquisition module, configured to acquire associated information of a user who uploaded the data to be classified in response to the classification confidence being greater than or equal to a first confidence threshold and less than or equal to a second confidence threshold; The first data classification module is configured to classify the data to be classified into the first data type in response to determining that the uploading user is a target user based on the association information, wherein the number of historical data of the first data type published by the target user within a historical period is greater than or equal to a quantity threshold.

9. An electronic device, comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 7.

10. A computer readable medium having computer program instructions stored thereon, wherein the computer program instructions can be executed by a processor to implement the method according to any one of claims 1 to 7.

11. A computer program product, comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 7.

Images