Sample data processing method and device, equipment, storage medium and program product

Abstract

The invention discloses a sample data processing method and device, equipment, a storage medium and a program product, and relates to the technical field of computers, in the method, an annotation file comprises result statistical areas in one-to-one correspondence with annotation results, each sample identifier is recorded in the result statistical areas of the annotation results needing to be associated, and the result statistical areas of the annotation results need to be associated. The sample identifier in each result statistical region is stored in a continuous storage region of the storage medium. Therefore, when the sample data with the target labeling result is screened, a plurality of sample identifiers associated with one labeling result can be read in one IO reading process, equivalently, the association relationship between a plurality of sample identifiers and labeling results can be read in one IO reading process, so that the IO frequency of data reading can be greatly reduced, and the data reading efficiency is improved. And the data reading efficiency is improved, so that the sample data processing efficiency is improved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a sample data processing method, apparatus, device, storage medium and program product. Background Art

[0002] Some artificial intelligence platforms support users to import sample data and process the sample data, such as modifying sample data of poor quality. At the same time, they also support users to create or import models to be trained and train the models based on the processed sample data.

[0003] At present, in order to ensure the quality of sample data in these artificial intelligence platforms, it is necessary to check the sample data one by one and manually modify the sample data of poor quality, and the processing efficiency of sample data is relatively low. Summary of the invention

[0004] The present application provides a sample data processing method, a sample data processing device, an electronic device, a non-volatile computer-readable storage medium, and a computer program product to at least solve the problem of low sample data processing efficiency in the related art.

[0005] The present application provides a sample data processing method, including: In response to receiving the first display instruction, display at least part of the sample data in the sample data set, the sample data set including a plurality of sample data, and the sample data having a sample identifier; In response to the displayed sample data being labeled, an association relationship between a sample identifier of the sample data and a labeling result is recorded in a labeling file, wherein the labeling result represents a data category to which the sample data belongs, and the labeling file includes a result statistical area corresponding to each of the labeling results. When recording the association relationship between the sample identifier and the labeling result, each sample identifier is recorded in a result statistical area of ​​the labeling result to be associated, and the sample identifier in each result statistical area is stored in a continuous storage area of ​​a storage medium; In response to receiving a screening instruction including a target annotation result, sample data having the target annotation result is screened in the sample data set according to the association between the sample identifier and the annotation result recorded in the annotation file, and the screened sample data is processed.

[0006] The present application also provides a sample data processing device, comprising: A display module, configured to display at least part of the sample data in the sample data set in response to receiving a first display instruction, wherein the sample data set includes a plurality of sample data and the sample data has a sample identifier; an association module, configured to record, in response to a labeling operation being performed on the displayed sample data, an association relationship between a sample identifier of the sample data and a labeling result in a labeling file, wherein the labeling result represents a data category to which the sample data belongs, and wherein the labeling file includes result statistical areas corresponding to the labeling results one by one, and when recording the association relationship between the sample identifier and the labeling result, each sample identifier is recorded in a result statistical area of ​​the labeling result to be associated, and the sample identifier in each result statistical area is stored in a continuous storage area of ​​a storage medium; A screening module is used to respond to receiving a screening instruction including a target annotation result, screen the sample data having the target annotation result in the sample data set according to the association relationship between the sample identifier and the annotation result recorded in the annotation file, and process the screened sample data.

[0007] The present application also provides an electronic device, comprising: a memory for storing a computer program; and a processor for implementing the steps of any of the above-mentioned sample data processing methods when executing the computer program.

[0008] The present application also provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned sample data processing methods are implemented.

[0009] The present application also provides a computer program product, including a computer program, which implements the steps of any of the above-mentioned sample data processing methods when executed by a processor.

[0010] Through this application, the annotation file includes a result statistics area that corresponds to the annotation result one by one, and each sample identifier is recorded in the result statistics area of ​​the annotation result to be associated, and the sample identifiers in each result statistics area are stored in a continuous storage area of ​​the storage medium. In this way, when screening sample data with target annotation results, multiple sample identifiers associated with a annotation result can be read in one IO reading process, which is equivalent to reading the association relationship between multiple sample identifiers and annotation results in one IO reading process. In this way, the number of IO times for data reading can be greatly reduced, the data reading efficiency can be improved, and the processing efficiency of sample data can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only 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 work.

[0012] Figure 1 It is a module diagram of some artificial intelligence platforms in technology; Figure 2 This is one of the example graphs showing sample data in pages; Figure 3 is a module schematic diagram of an artificial intelligence platform provided in some embodiments of the present application; Figure 4 is a schematic diagram of a sample data annotation interface provided in some embodiments of the present application; Figure 5 is a schematic diagram of a sample data processing interface provided by some embodiments of the present application; Figure 6 is a schematic diagram of a sample data processing interface provided by other embodiments of the present application; Figure 7 is a schematic diagram of a sample data processing interface provided by other embodiments of the present application; Figure 8 is a schematic diagram of a sample data processing interface provided by other embodiments of the present application; Fig. 9 is a schematic diagram of a sample data processing interface provided by other embodiments of the present application; Fig.10 is a schematic diagram of a sample data processing interface provided by other embodiments of the present application; Fig.11 is a schematic diagram of a sample data processing interface provided by other embodiments of the present application; Fig.12 is a schematic diagram of a sample data processing interface provided by other embodiments of the present application; Fig.13 is a schematic diagram of a summary browsing interface of annotation results provided in some embodiments of the present application; Fig.14 is a flowchart of a sample data processing method provided by some embodiments of the present application; Fig.15 is a module schematic diagram of a sample data processing device provided in some embodiments of the present application; Fig.16 It is a schematic diagram of the structure of an electronic device provided in some embodiments of the present application. DETAILED DESCRIPTION

[0013] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only 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.

[0014] It should be noted that, in the description of this application, the terms "include", "comprise" or any other variant 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 includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. The terms "first", "second", etc. in this application are used to distinguish similar objects, and are not used to describe a specific order or sequence.

[0015] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0016] Combined with reference Figure 1 , which is a module diagram of the artificial intelligence platform in some technologies. Figure 1 In the AI ​​platform, the AI ​​platform includes a sample data processing system and a model training system. The sample data processing system can be used to process sample data input by the user and provide the processed sample data to the model training system. Based on the sample data provided by the sample data processing system, the model training system can train or fine-tune the model created or imported by the user.

[0017] The sample data processing system may include an initial data management module and a data set management module. The user may store the sample data in the initial data management module by locally uploading, remotely downloading, etc. Alternatively, the output data of the data processing task may be used as sample data and stored in the initial data management module. The initial data management module may include a storage medium for storing the sample data.

[0018] Among the sample data stored in the initial data management module, the user can select at least part of the sample data as a sample data set and import the sample data set into the data set management module. The data set management module may include a data processing interface. The data processing interface can be used to display at least part of the sample data in the sample data set. In the data processing interface, the user can process the sample data, for example, modify sample data of poor quality. The processed sample data can be saved in the cache of the model training system. In this way, the model training system can perform model training or fine-tuning based on the sample data in the cache.

[0019] The sample data may be text data, and the unit of measurement may be "pieces". The sample data may be different for different model training or fine-tuning tasks. For example, for the training or fine-tuning task of the question-answering model, each piece of sample data may include a question and an answer, and the questions and answers in different pieces of sample data may be different. The specific examples are shown in Table 1.

[0020] Table 1 Sample data

[0021] In the above Table 1, the question and answer in the second row can be used as a sample data, and the question and answer in the third row can be used as another sample data. After the question-answering model is trained or fine-tuned using sample data similar to that in Table 1, the question-answering model can answer questions raised by users.

[0022] For another example, for the training or fine-tuning task of a recommendation model, each piece of sample data may include the user's historical behavior, user characteristics, item characteristics, and interaction results. The historical behavior, user characteristics, item characteristics, and interaction results in different pieces of sample data are not exactly the same. Specific examples are shown in Table 2.

[0023] Table 2 Sample data

[0024] In the above Table 2, the data in the first row, the second row, and the third row can be used as a piece of sample data respectively. After training or fine-tuning the recommendation model using sample data similar to that in Table 2, the recommendation model can learn the potential relationship between user behavior and interaction results, such as the potential relationship between user age and movie ID watched, and the potential relationship between user ID and article ID read and reading time. In this way, after the recommendation model is trained, the user ID, user age and other information are input into the recommendation model, and the recommendation model can recommend movie IDs and article IDs that match the user ID.

[0025] The sample data may be stored in a sample file. Each sample file may include multiple pieces of sample data, for example, each sample file may include 10,000 or 20,000 pieces of sample data. In the storage medium, the sample data in all sample files are stored in an aggregated manner, and each sample data has a corresponding row number, which may be used as a sample identifier of the sample data. For example, Table 3 exemplifies the correspondence between sample data and sample identifiers: Table 3 Correspondence between sample identifier and sample data

[0026] The user can select one or more sample files in the initial data management module, and the sample data in the selected sample files constitute the sample data set. When there are many sample data in the sample data set, the data set management module can display the sample data in pages in the data processing interface. The so-called page display means that the sample data in the sample data set is divided into multiple subsets, and the sample data in one subset is displayed on each page. For ease of understanding, please refer to Figure 2 , which is one of the example graphs showing sample data in pages. Figure 2 In the example, assuming that the data processing interface supports displaying 3 pieces of sample data at a time, and the sample data set includes 300 pieces of sample data, the sample data in the sample data set can be divided into 100 subsets and each subset includes 3 pieces of sample data, and the sample data in each subset is displayed as one page.

[0027] In the data processing interface, users can check each sample data in turn according to the target conditions that the model training data needs to meet, and determine whether the sample data meets the target conditions. Among them, the target conditions include but are not limited to high data accuracy, representative data and good timeliness, complete data and low noise, etc. For any sample data, the data quality of the sample data can be determined according to the number of target conditions that the sample data meets. For example, if the sample data meets all the target conditions, it can be considered that the data quality of the sample data is relatively good; if the sample data meets some of the target conditions, it can be considered that the data quality of the sample data is medium; if the sample data does not meet the target conditions, it can be considered that the data quality of the sample data is relatively poor. According to the data quality of the sample data, users can process the sample data in the data processing interface, such as modifying the sample data with poor data quality to meet the target conditions. In this way, the sample data after processing can have a higher quality, thereby ensuring the accuracy of subsequent model training. These technologies require users to check and modify sample data one by one, which is a large workload and is not applicable.

[0028] In view of this, this application Figure 1 The artificial intelligence platform shown has been improved. Figure 3 , which is a module diagram of the artificial intelligence platform provided in some embodiments of the present application. Figure 3 and Figure 1Basically similar, the main difference is: the data annotation, data filtering and data export functions are newly added in the data set management module. Based on these newly added functions, users can annotate the data quality of sample data in the data set management module. The annotation results can characterize the data category to which the sample data belongs, and the data category can be divided according to the quality of the sample data. For example, sample data with good data quality can be used as the first data category, and the sample data with good data quality can be annotated as "good"; sample data with medium data quality can be used as the second data category, and the sample data with medium data quality can be annotated as "medium"; sample data with poor data quality can be used as the third data category, and the sample data with poor data quality can be annotated as "poor". Based on the annotation results, sample data that meets the data quality requirements can be selected as training data for the model. For example, in the sample data set that has been annotated, sample data marked as "good" are selected as training data for the model. In this way, sample data with poor or medium data quality in the sample data set can be directly eliminated without manually modifying sample data with poor data quality, thereby reducing the workload.

[0029] To realize the above-mentioned newly added functions, the data set management module of the present application may specifically include a sample data annotation interface and a sample data processing interface, which are described below respectively.

[0030] Combined with reference Figure 4 , which is a schematic diagram of a sample data annotation interface provided in some embodiments of the present application. Figure 4 It includes a sample data display area 41, a sample annotation area 42 and a sample data switching area 43. In the sample data display area 41, one of the sample data in the sample data set can be displayed. In the sample annotation area 42, multiple data category options can be displayed, such as "good", "medium", and "poor".

[0031] By selecting an option in the sample annotation area 42, the user can annotate the sample data displayed in the sample data display area 41 to obtain the annotation result of the sample data. Figure 4 In the example, the option “good” in the sample annotation area 42 is selected, which means that the sample data displayed in the sample data display area 41 is marked as “good”, that is, the marking result of the sample data is “good”.

[0032] When the sample data in the sample data display area 41 has been annotated, the user can update the annotation result of the sample data by changing the selected option in the sample annotation area 42. For example, after the selected option in the sample annotation area 42 is changed from "good" to "medium", it means that the annotation result of the sample data is updated from "good" to "medium".

[0033] After the sample data in the sample data display area 41 are labeled, the sample data displayed in the sample data display area 41 can be switched based on the first switching component 431 or the second switching component 432 in the sample data switching area 43. For example, it is assumed that the sample data currently displayed in the sample data display area 41 is sample data A1. After the sample data A1 is labeled, the first switching component 431 is clicked, and the sample data display area 41 can display the previous sample data of the sample data A1, and the second switching component 431 is clicked, and the sample data display area 41 can display the next sample data of the sample data A1.

[0034] After the user completes labeling of any sample data displayed in the sample labeling area 42, the data set management module can record the association between the sample identifier of the sample data and the labeling result. For example, suppose the sample identifier of sample data A2 is "201". After the user labels sample data A2 as "medium" in the sample labeling area 42, the data set management module can record the association between the sample identifier "201" and the labeling result "medium".

[0035] Based on the association between sample identification and annotation results, Figure 5 The sample data processing interface shown displays each sample data and the annotation results of the sample data. For example, the display of sample data A1 and the annotation results of sample data A1 is used as an example for explanation. Sample data A1 and the sample identifier 200 of sample data A1 can be read from the sample data set first. According to the sample identifier 200, the annotation result corresponding to the sample identifier 200 can be searched in the association relationship between the sample identifier and the annotation result, and the annotation results corresponding to the sample data A1 and the sample identifier 200 can be displayed simultaneously in the sample data processing interface. In this way, the user can browse the annotation results of each sample data one by one.

[0036] Combined with reference Figure 6 . In some embodiments, the sample data processing interface also includes a filtering component for the annotation results, that is, the component in the first dotted box 61. The filtering component can be a component in the form of a drop-down list. In the drop-down list of the filtering component, options representing various data categories can be included, such as "good", "medium", and "poor". At the same time, options representing all annotation results can also be included, such as "all". The user can select one of the options in the drop-down list of the filtering component according to actual needs. According to the selected option, only sample data with corresponding annotation results can be displayed in the sample data processing interface. For example Figure 6 In the filter component drop-down list, if the option "Good" is selected, only sample data with the marked result of "Good" can be displayed in the sample data processing interface. Figure 7In the sample data processing interface shown, in the drop-down list of the filter component, the option "All" is selected, indicating that all the marked sample data are displayed. Figure 6 and Figure 7 It can be seen that after setting the filtering component, users can browse sample data with target annotation results according to actual needs, which is more applicable.

[0037] Combined with reference Figure 8 In some embodiments, the sample data processing interface further includes an export component, namely, the component within the second dashed box 81. Figure 8 In the sample data processing interface shown, based on the data filtering function of the filtering component, only sample data with target annotation results can be displayed. When the export component is triggered, these sample data with target annotation results can be exported, and the exported sample data can be used as a new sample data set. The new sample data set can only include sample data with target annotation results. For example, assuming that the option "good" is selected in the drop-down list of the filtering component, only sample data with an annotation result of "good" can be displayed in the sample data processing interface. At this time, if the export component is triggered, the sample data with an annotation result of "good" can be exported to obtain a new sample data set. In the new sample data set, only sample data with an annotation result of "good" is included, and sample data with poor and medium quality in the original sample data set are eliminated. Therefore, the new sample data set can be used for model training or fine-tuning. In the scheme of the present application, based on the method of quality annotation, sample data with better quality can be screened from the sample data set as sample data for model training or fine-tuning, without the user having to modify the sample data with poor quality one by one, which greatly reduces the workload.

[0038] In summary, in the artificial intelligence platform of this application, users can obtain new sample data sets for model training based on the following steps: 1) In the initial data management module, select the required sample files and import them into the data set management module to obtain the sample data set; 2) In Figure 4 In the sample data annotation interface shown, each sample data in the sample data set is annotated to obtain the correlation relationship between the sample identifier of each sample data and the annotation result; 3) Based on the relationship between sample identification and annotation results, Figure 8 In the sample data processing interface shown, by selecting the target annotation result (such as "good") through the filter component, only the sample data with the target annotation result can be displayed in the sample data processing interface. At this time, the export component is triggered to export the sample data with the target annotation result. In this way, a new sample data set for model training is obtained.

[0039] Combined with reference Fig. 9 In some embodiments, each sample data in the sample data processing interface also has a corresponding annotation component, i.e., the component in the third dotted box 91. For any sample data, if the annotation component corresponding to the sample data is triggered, the sample data annotation interface for the sample data can be displayed. The user can update the annotation results of the sample data in the sample data annotation interface of the sample data. In this way, when the user browses the annotation results of the sample data in the sample data processing interface, if the annotation results of the sample data are found to be incorrect, the annotation results of the sample data can be modified in a timely and convenient manner.

[0040] Of course, when the sample data is not labeled, you can also trigger the labeling of the sample data through the sample data processing interface. For example, Fig.10 In the example, some sample data are not labeled. In this case, the labeling of each sample data can be triggered based on the labeling component corresponding to the sample data.

[0041] Continue reading Fig.10 In the sample data processing interface, if multiple unlabeled sample data can have the same labeling result, such as "good", in this case, if the labeling operation is performed on the multiple sample data separately, it will inevitably cause cumbersome operations. In view of this, combined with the reference Fig.11 In some embodiments, the sample data processing interface further includes a batch annotation component, i.e., the component in the fourth dotted box 111. If multiple unannotated sample data can have the same annotation result, the multiple sample data can be selected and the batch annotation component can be triggered. At this time, the sample data annotation interface for the multiple sample data can be displayed. In the sample data annotation interface, the multiple sample data can be batch annotated. In this way, the annotation process can be greatly simplified and the annotation efficiency can be improved.

[0042] Of course, it is understandable that Fig.11 If you need to update the annotation results of multiple sample data, and the updated annotation results of the multiple sample data are the same, you can also follow Fig.10 The method is used to select the multiple sample data and trigger the batch annotation component. In this way, the annotation results of the multiple sample data can be updated in batches in the displayed sample data annotation interface, thereby improving the updating efficiency of the annotation results.

[0043] Combined with reference Fig.12 In some embodiments, the sample data processing interface further includes a progress viewing component, namely, the component within the fifth dotted frame 121. When the progress viewing component is triggered, the following may be displayed: Fig.13The summary browsing interface of the annotation results shown. In the summary browsing interface of the annotation results, the total number of sample data in the sample data set and the number of labeled sample data can be displayed. At the same time, the number of sample data corresponding to each annotation result can also be displayed. In this way, it is convenient for users to understand the summary of sample data annotation.

[0044] In summary, based on the sample data annotation scheme of the present application, sample data with better quality can be screened out from the sample data set as model training data, thereby avoiding the problem of heavy workload caused by modifying the sample data one by one. However, in this solution based on sample data annotation, it is necessary to store the association between the sample identifier and the annotation result. At present, in some technologies, the association between the sample identifier and the annotation result of each sample data can be stored as one of the records in the database, as shown in Table 4: Table 4 Correlation between sample identification and annotation results

[0045] The association between sample identification and labeling results can be stored in the database in a row storage manner. Row storage means that the data in each record of the database is stored in a continuous storage area of ​​the storage medium, and the data of different records can be stored in discontinuous storage areas of the storage medium. This storage method has the following two problems: 1) Since the number of sample data is usually large (e.g., millions of records), there are also many records of the association between sample identifiers and annotation results in the database, which will cause great performance pressure on the database; 2) When screening sample data with target annotation results, it is necessary to read records in the database one by one and search for records with target annotation results. Since data of different records can be stored in discontinuous storage areas of the storage medium, and each storage area requires at least one IO read, which is equivalent to at least one IO read for each data record, when screening sample data with target annotation results, the number of IO (Input / Output) reads required is also relatively large, which greatly reduces the data reading efficiency and further reduces the processing efficiency of sample data.

[0046] In view of this, the present application provides a sample data processing method, which can improve the processing efficiency of sample data. The sample data processing method can be applied to Figure 3 The sample data processing system in the embodiment of the present invention, or an electronic device applicable to running the sample data processing system. The electronic device may include but is not limited to a tablet computer, a laptop computer, a desktop computer, a server, etc. Fig.14 , which is a flow chart of a sample data processing method provided in some embodiments of the present application. Fig.14 In the method, the sample data processing method comprises the following steps: Step S141: in response to receiving a first display instruction, display at least part of sample data in a sample data set, where the sample data set includes a plurality of sample data, and the sample data has a sample identifier.

[0047] Specifically, the sample data set refers to a set of sample data imported from the initial data management module to the data set management module. Before executing step S141, the sample data processing system may import the selected sample data into the data set management module in response to one or more sample data in the initial data management module being selected, as sample data in the sample data set. It should be noted that when the sample data is stored in a sample file, the sample data processing system may import the selected sample file into the data set management module in response to one or more sample texts in the initial data management module being selected, and use the sample data in the selected sample file as sample data in the sample data set.

[0048] Furthermore, when the selected sample data is imported into the data set management module, if the sample data in the sample data set is stored in the initial data management module and the data set management module at the same time, it will inevitably cause duplicate storage of the sample data, thereby causing a waste of storage space. In view of this, in this embodiment, the storage path of the selected sample data in the initial data management module can be imported into the data set management module. When the data set management module displays at least part of the sample data in the sample data set, it can obtain the sample data to be displayed from the initial data management module based on the storage path. That is, the entity data of the sample data is stored in the initial data management module, and there is only the storage path of the sample data in the data set management module, but no entity data of the sample data. When the data set management module needs to display the sample data, it can obtain the sample data from the initial data management module based on the storage path. In this way, the problem of wasting storage space caused by storing the same sample data in the initial data management module and the data set management module at the same time can be avoided.

[0049] Furthermore, after the sample data set is imported into the data set management module, if the sample data in the sample data set is modified in the initial data management module, it may cause inaccurate labeling results in the data set management module. For example, after the sample data A3 is labeled as "good" in the data set management module, if the word order, content, etc. of the sample data A3 are adjusted in the initial data management module, making the quality of the sample data A3 worse, then in the data set management module, the labeling results of the sample data A3 will be inaccurate.

[0050] In this embodiment, in order to avoid the problem of inaccurate labeling results caused by changes in sample data in the initial data management module, after the storage path is imported into the data set management module, the selected sample data can be identified in the initial data management module to distinguish the selected sample data and the unselected sample data in the initial data management module. Among them, the selected sample data refers to the sample data imported into the data set management module, and the unselected sample data refers to the sample data that has not been imported into the data set management module. The identification method of the sample data can be selected according to actual needs, and this application does not limit this. For example, the selected sample data can be identified by an icon, and the unselected sample data can be identified without an icon. In this way, the user can determine whether the sample data has been imported into the data set management module based on the icon. For the sample data that has been imported into the data set management module, it can be not modified in the initial data management module, thereby ensuring the accuracy of the labeling results in the data set management module.

[0051] Step S142, in response to the displayed sample data being labeled, the association relationship between the sample identifier of the sample data and the labeling result is recorded in the labeling file, the labeling result represents the data category to which the sample data belongs, the labeling file includes a result statistical area corresponding to the labeling result one by one, when recording the association relationship between the sample identifier and the labeling result, each sample identifier is recorded in the result statistical area of ​​the labeling result to be associated, and the sample identifier in each result statistical area is stored in a continuous storage area of ​​the storage medium.

[0052] For the relationship between labeling results, data categories, sample identifiers and labeling results, please refer to the above description and will not be repeated here.

[0053] In this embodiment, on the one hand, the association relationship between the sample identifier and the annotation result is saved in the annotation file, and the data in the annotation file can be directly written to the storage medium. In this way, it is possible to avoid saving the association relationship between the sample identifier and the annotation result to the database and causing excessive performance pressure on the database. On the other hand, the annotation file includes a result statistics area that corresponds to the annotation result one by one, and each sample identifier is recorded in the result statistics area of ​​the annotation result to be associated, and the sample identifier in each result statistics area is saved in a continuous storage area of ​​the storage medium. In this way, multiple sample identifiers associated with a annotation result can be read in one IO reading process, which is equivalent to reading multiple association relationships between sample identifiers and annotation results in one IO reading process. In this way, the number of IO times for data reading can be greatly reduced, thereby improving the data reading efficiency and the processing efficiency of sample data.

[0054] For ease of understanding, take the labeling results of "good", "medium", and "poor" as an example. The result statistical area division in the labeling file can be similar to the following: Streams: { "outs":[ { "result": "good", "ID": "0, 3, 4, 5, 7,..., 20", }, { "result":"Medium", 〝ID〞:〝1、2〞, }, { "result": "poor", 〝ID〞:〝6〞, } ] } In the above example, result indicates the labeling result, ID indicates the sample ID, and the area where each result is located can be used as a result statistics area. The labeling result "good" is associated with sample identifiers 0, 3, 4, 5, 7, ..., 20, indicating that the labeling result corresponding to the sample data with sample identifiers 0, 3, 4, 5, 7, ..., 20 is "good"; The labeling result "medium" is associated with sample identifiers 1 and 2, indicating that the labeling result corresponding to the sample data with sample identifiers 1 and 2 is "medium"; The labeling result "bad" is associated with the sample identifier 6, indicating that the labeling result corresponding to the sample data with the sample identifier 6 is "bad".

[0055] According to the above example, when storing data in a storage medium, the labeling result "good" and sample identification 0, 3, 4, 5, 7, ..., 20 can exist in a continuous storage area, and the labeling result "medium" and sample identification 1, 2 can exist in another continuous storage area. And so on. In this way, when screening sample data with target labeling results, at least one IO read is required to read all the associations between sample identification and labeling results. Compared with some technologies that save the associations between sample identification and labeling results in a database, the number of IO reads required for some technologies may be millions of times. The technical solution of the present application can greatly reduce the number of IO reads, thereby improving the processing efficiency of sample data.

[0056] Step S143, in response to receiving a screening instruction including a target annotation result, screening sample data having the target annotation result in the sample data set according to the association relationship between the sample identifier and the annotation result recorded in the annotation file, and processing the screened sample data.

[0057] Specifically, regarding the selection of sample data with target annotation results, please refer to the above steps S142 and Figure 5 , Figure 6 The relevant description is not repeated here.

[0058] Process the sample data obtained by screening, including but not limited to displaying the sample data with target annotation results (such as Figure 5 As shown), update the annotations of the displayed sample data (as shown Fig. 9 , 10 As shown in ), export the sample data with target annotation results as a new sample dataset (as shown in Figure 8 shown) etc.

[0059] In summary, in the technical solutions of some embodiments of the present application, the annotation file includes a result statistics area corresponding to the annotation results one by one, and each sample identifier is recorded in the result statistics area of ​​the annotation result to be associated, and the sample identifiers in each result statistics area are stored in a continuous storage area of ​​the storage medium. In this way, when screening sample data with target annotation results, multiple sample identifiers associated with a annotation result can be read in one IO reading process, which is equivalent to reading the association relationship between multiple sample identifiers and annotation results in one IO reading process. In this way, the number of IO times for data reading can be greatly reduced, the data reading efficiency can be improved, and the processing efficiency of sample data can be improved.

[0060] Furthermore, referring to the example of result statistical area division in the above annotation file, it can be understood that when there are many records of the association relationship between sample identifiers and annotation results (for example, millions of records), there will be many sample identifiers in at least some result statistical areas. In this way, when reading the sample identifier in a single result statistical area, the duration of a single IO read may be very long, which will also reduce the data reading efficiency and thus affect the processing efficiency of the sample data.

[0061] In view of this, in some embodiments, in the sample data set, the sample identification of the sample data can be divided into multiple sample identification intervals. For example, sample identifications 1 to 1000 are divided into the first sample identification interval, and sample identifications 1001 to 2000 are divided into the second sample identification interval. And so on.

[0062] When dividing the sample identification intervals, the number of sample identifications in each sample identification interval may be the same, or the number of sample identifications in at least some sample identification intervals may be different. For example, each sample identification interval may include 1000 sample identifications, or the first sample identification interval may include 500 sample identifications, and the second sample identification interval may include 1000 sample identifications. This application does not limit this.

[0063] The annotation file may also include interval statistical areas corresponding to the sample identification intervals, and each interval statistical area includes result statistical areas corresponding to the annotation results. For ease of understanding, assuming that the annotation results are "good", "medium", and "poor", and in the order of the size of the sample identifications, every 1,000 sample identifications are considered as a sample identification interval, then the division of the interval statistical area and result statistical area in the annotation file may be similar to the following: Streams: { "index":0, "outs":[ { "index":1, "result": "good", "ID": "0, 3, 4, 5, 7,..., 20", }, { "result":"Medium", 〝ID〞:〝1、2〞, }, { "result": "poor", 〝ID〞:〝6〞, } ] }, { "index":1, "outs":[ { "result": "good", "ID": "21,...,24,27,...,40", }, { "result":"Medium", 〝ID〞:〝25〞, }, { "result": "poor", 〝ID〞:〝26〞, } ] } In the above example, index represents the sample identification interval, result represents the labeling result, and ID represents the sample identification. The area where each index is located represents an interval statistical area. For example, the interval statistical area "index":0 can correspond to the sample identification interval 1~1000, and the interval statistical area "index":1 can correspond to the sample identification interval 1001~2000.

[0064] Based on the above description, the association relationship between the sample identifier and the annotation result of the sample data recorded in the annotation file may include: Determine the target sample identification interval where the sample identification is located according to the divided sample identification interval; In the interval statistics area of ​​the target sample identification interval, the sample identification is recorded in the result statistics area of ​​the annotation result to be associated.

[0065] For example, suppose the sample identifier of sample data A1 is 1005, and the annotation result is "good". When recording the association between the sample identifier of sample data A1 and the annotation result in the annotation file, it can be first determined that the sample identifier 1005 is located in the sample identifier interval 1001~2000. Therefore, the result statistical area with the annotation result "good" can be searched in the interval statistical area corresponding to the sample identifier interval 1001~2000, and the sample identifier 1005 can be recorded in the found result statistical area. In this way, the amount of data read in a single IO can be reduced, thereby improving data reading efficiency.

[0066] In some embodiments, the interval statistical area can be dynamically divided, that is, when there is annotated sample data in a sample identification interval, it will trigger the division of the interval statistical area corresponding to the sample identification interval in the annotation file. In this way, it is possible to avoid dividing redundant and unnecessary interval statistical areas in the annotation file. In view of this, before recording the association between the sample identification and the annotation result of the sample data in the interval statistical area of ​​the target sample identification interval, it is also possible to check whether the interval statistical area of ​​the target sample identification interval is divided in the annotation file. If not, the interval statistical area of ​​the target sample identification interval is divided in the annotation file, and the result statistical area corresponding to the annotation result is divided in the interval statistical area of ​​the target sample identification interval.

[0067] Further, based on Fig. 9From the relevant description, it can be known that after the labeling of the sample data is completed, the labeling result of the sample data can also be modified. In this case, the sample identification and labeling result of the same sample data may be repeatedly written into the labeling file, which may cause data conflict problems. For example, suppose that in the sample data labeling interface, sample data A5 is labeled as "good". At this time, an association relationship will be established between the sample identification of sample data A5 and the labeling result "good" in the labeling file. After the labeling of sample data A5 is completed, if the labeling result of sample data A5 is modified to "medium" in the sample data labeling interface, then an association relationship will be established between the sample identification of sample data A5 and the labeling result "medium" in the labeling file. In this way, in the labeling file, the sample identification of sample data A5 will be associated with the two labeling results, which will cause data conflict problems.

[0068] In view of this, in some embodiments, when recording the association relationship between the sample identifier of the sample data and the annotation result in the annotation file, the method of the present application may further include: If there is a target sample identifier in the annotation file that is the same as the sample identifier of the sample data, and the annotation result associated with the target sample identifier is different from the annotation result of the sample data, the target sample identifier is deleted from the annotation file, and the association between the sample identifier of the sample data and the annotation result is recorded in the annotation file.

[0069] For example, take the above-mentioned sample data A5 as an example. When recording the association relationship between the sample identifier of sample data A5 and the annotation result "medium" in the annotation file, if the sample identifier of sample data A5 (i.e., the target sample identifier) ​​already exists in the annotation file, and in the annotation file, the sample identifier of sample data A5 is associated with the annotation result "good", then the sample identifier of sample data A5 can be deleted in the annotation file, and the association relationship between the sample identifier of sample data A5 and the annotation result "medium" can be recorded in the annotation file. In this way, on the one hand, when the user updates the annotation result of the sample data in the sample data annotation interface, the annotation result of the sample data can be synchronously updated in the annotation file to ensure the accuracy of the annotation result in the annotation file; on the other hand, it can avoid associating the sample identifier of the same sample data with different annotation results in the annotation file, thereby causing data conflicts.

[0070] Furthermore, when the labeling result associated with the target sample identifier is the same as the labeling result of the sample data, although it will not cause data conflict problems, it may cause data duplication problems. For example, suppose that in the sample data labeling interface, sample data A6 is labeled as "good", and the sample identifier of sample data A6 is associated with the labeling result "good" in the labeling file. If the labeling result of sample data A6 is updated in the sample data labeling interface, and the updated labeling result is still "good", then the sample identifier of sample data A6 will be associated with the labeling result "good" again in the labeling file. In this way, there will be two sample identifiers of sample data A6 in the labeling file, and these two sample identifiers correspond to the labeling result "good", which causes the problem of sample identifier duplication.

[0071] In view of this, in some embodiments, if the target sample identifier exists in the annotation file, and the annotation result associated with the target sample identifier is the same as the annotation result of the sample data, the target sample identifier is retained in the annotation file, and the association between the sample identifier and the annotation result of the sample data is stopped in the annotation file. In this way, on the one hand, the problem of duplicate sample identifiers in the annotation file can be avoided, and on the other hand, the IO consumption when writing data to the annotation file can be reduced.

[0072] In some embodiments, the annotation file may be triggered to be created when the sample data in the sample data set is annotated for the first time. That is, when the user annotates the sample data in the sample data set for the first time in the sample data annotation interface, the creation of the annotation file corresponding to the sample data set is triggered. In view of this, in the above step S142, before writing the sample identification and annotation results of the sample data into the annotation file, it can also be determined whether the sample data is the first sample data to be annotated in the sample data set. If so, the creation of the annotation file is triggered, and the sample identification and annotation results of the sample data are written into the newly created annotation file. If not, the association relationship between the sample identification and annotation results of the sample data can be recorded in the annotation file that has been created.

[0073] In some embodiments, after the annotation file is created, the annotation file is also set to a hidden state. In this way, on the one hand, the content in the annotation file can be prevented from being accidentally modified by other users, and on the other hand, the annotation file can be prevented from being accidentally deleted by other users.

[0074] In some embodiments, when determining whether the sample data is the first labeled sample data in the sample data set, it may be determined whether there is a labeling file. If not, it is determined that the sample data is the first labeled sample data in the sample data set.

[0075] In other embodiments, a labeling state parameter may be set for the sample data set. When the value of the labeling state parameter is a first value, it indicates that the sample data of the sample data set is labeled for the first time, and when the value of the labeling state parameter is a second value, it indicates that the sample data of the sample data set is not labeled for the first time. Before recording the association relationship between the sample identifier and the labeling result of the sample data each time, it may be determined whether the value of the labeling state parameter corresponding to the sample data set is the first value. If so, it is determined that the sample data is the first labeled sample data in the sample data set, and the value of the labeling state parameter is changed to the second value.

[0076] In some embodiments, when displaying sample data in the sample data set, the data set management module may also stop displaying the sample data in response to receiving a first exit instruction. For example, when a user exits the sample data processing interface or the sample data annotation interface, the first exit instruction may be triggered to cause the data set management module to stop displaying the sample data. After stopping displaying the sample data, the data set management module may also display at least part of the sample data in the sample data set again in response to receiving a second display instruction. For example, when a user logs back into the sample data processing interface or the sample data annotation interface, the second display instruction may be triggered to cause the data set management module to display at least part of the sample data in the sample data set again.

[0077] In some embodiments, after the first user who annotates the sample data exits the sample data processing interface or the sample data annotation interface, other second users may modify the sample data in the sample data set, such as modifying the sentence structure of the sample data. In this case, if the first user has already annotated the sample data in the sample data set, the operation of modifying the sample data by the second user may affect the accuracy of the annotation result. For example, after the first user exits the sample data processing interface or the sample data annotation interface, the second user modifies the sample data A8 marked as "good", causing the data quality of the sample data A8 to decline. In this case, the annotation result of the sample data A8 may no longer be accurate, but the first user is not aware of the risk of inaccurate annotation result of the sample data A8.

[0078] In view of this, when the second display instruction is received, it can be determined whether the sample data in the sample data set is updated after the display of the sample data is stopped. If the sample data in the sample data set is updated after the display of the sample data is stopped, then when at least part of the sample data in the sample data set is displayed again, an update prompt of the sample data is displayed at the same time, for example, a pop-up prompt is displayed in the sample data processing interface. In this way, based on the update prompt, the first user can know that other users have modified the sample data after he exits the sample data processing interface or the sample data annotation interface, and then the accuracy of the existing annotation results can be checked. In this way, the accuracy of the annotation results can be guaranteed.

[0079] In some embodiments, when the first user exits the sample data processing interface or the sample data annotation interface, a first hash value of the sample data set can be calculated. When the first user logs into the sample data processing interface or the sample data annotation interface again, a second hash value of the sample data set can be calculated. If the first hash value and the second hash value are different, it means that after the first user exits the sample data processing interface or the sample data annotation interface, a second user modifies the sample data; if the first hash value and the second hash value are the same, it means that after the first user exits the sample data processing interface or the sample data annotation interface, no second user modifies the sample data. However, this comparison method based on hash values ​​requires traversing the sample data in the sample data set, and the amount of data calculation is relatively large.

[0080] In view of this, in some other embodiments, whether the sample data in the sample data set is updated may be determined based on the following method: When the display of the sample data stops, the first data modification time of the sample data set is recorded in the annotation file, where the first data modification time is the last modification time of the sample data set when the display of the sample data stops; When displaying at least part of the sample data in the sample data set again, obtaining a second data modification time of the sample data set, where the second data modification time is the last modification time of the sample data set when the second display instruction is received; If the first data modification time and the second data modification time are different, it means that the sample data in the sample data set is updated after the display of the sample data is stopped.

[0081] Specifically, the last modification time may be attribute information of the sample data set. When the sample data in the sample data set changes, the last modification time of the sample data set will change synchronously. Based on this, the above-mentioned first data modification time indicates the last modification time of the sample data set when the first user exits the sample data processing interface or the sample data annotation interface. The second data modification time indicates the last modification time of the sample data set when the first user logs into the sample data processing interface or the sample data annotation interface again. If the first data modification time and the second data modification time are not the same, it means that after the first user exits the sample data processing interface or the sample data annotation interface, the sample data in the sample data set has changed, that is, it has been modified by the second user. If the first data modification time and the second data modification time are the same, it means that after the first user exits the sample data processing interface or the sample data annotation interface, the sample data in the sample data set has not changed, that is, it has not been modified by the second user.

[0082] In the above embodiment, whether the sample data in the sample data set is updated is determined based on the last modification time of the sample data set, without traversing the sample data in the sample data set, thereby reducing the amount of calculation.

[0083] Specifically, the first data modification time can be recorded in the file name of the annotation file. For example, the file name of the annotation file can be in the form of "annotation file A+last modification time". In this way, when reading the association between the sample identifier and the annotation result from the annotation file, the latest last modification time of the sample data set can be obtained based on the file name of the annotation file, which is more convenient.

[0084] In some embodiments, the method of the present disclosure further comprises: In response to receiving the second exit instruction, stopping displaying the sample data; After the display of sample data stops, if the sample data in the sample data set is updated, the changed target sample data is recorded; In response to receiving the third display instruction, at least part of the sample data in the sample data set is displayed again, and at the same time, an update prompt of the target sample data is displayed.

[0085] Specifically, when displaying the update prompt of the target sample data, the sample identifier of the target sample data may be displayed in a pop-up window, or the font, color, etc. of the target sample data may be changed. In this way, the first user can distinguish the modified target sample data from other unmodified sample data according to the update prompt, and then check the annotation result of the target sample data to ensure the accuracy of the annotation result.

[0086] In the above embodiment, by recording the changed target sample data, an update prompt for the target sample data can be directly displayed, so that the first user does not need to check all the annotation results of the sample data, which can greatly reduce the workload.

[0087] In some embodiments, the above-mentioned simultaneous display of update prompts of target sample data includes: According to the sample identifier of the target sample data, the association relationship between the sample identifier recorded in the annotation file and the annotation result, determining whether the target sample data has an annotation result; If the target sample data has a labeled result, the update prompt of the target sample data will be displayed at the same time; If the target sample data does not have a labeling result, the update prompt of the target sample data will not be displayed.

[0088] Specifically, in these embodiments, after obtaining the changed target sample data, it is determined whether the target sample data is labeled. If the target sample data is not labeled, the second user's operation of modifying the target sample data will not affect the accuracy of the labeled result, so no update prompt is required; if the target sample data is labeled, the second user's operation of modifying the target sample data will affect the accuracy of the labeled result, so an update prompt is required. In this way, on the one hand, the number of update prompts can be reduced, and on the other hand, the first user can be prevented from performing invalid accuracy checks on the labeled results of the sample data.

[0089] In some embodiments, after labeling the sample data in the sample data set, the method of the present application further includes: In response to receiving the statistical instruction, based on the association between the sample identification subgroups and the annotation results recorded in the annotation file, counting the number of sample identifications corresponding to each annotation result; Display the number of sample identifiers corresponding to each annotation result.

[0090] For details, please refer to Fig.13 The relevant description is not repeated here.

[0091] In some embodiments, the sample data with target annotation results are also used for page display in the sample data processing interface; the above-mentioned screening of sample data with target annotation results in the sample data set includes: According to the number of sample data to be displayed on the current page, a corresponding number of sample data with target annotation results are filtered in the sample data set.

[0092] That is, when displaying sample data, you can obtain sample data with target annotation results in batches according to the number of sample data that can be displayed on the sample data processing interface, without having to obtain all the sample data with target annotation results. In this way, you can reduce the amount of data obtained at a time and improve data processing efficiency.

[0093] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment 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.

[0094] Combined with reference Fig.15 , a sample data processing device is provided for some embodiments of the present application. Fig.15 In the example, the sample data processing device comprises: A display module 151, configured to display at least part of the sample data in the sample data set in response to receiving the first display instruction, where the sample data set includes a plurality of sample data and the sample data has a sample identifier; An association module 152 is used for recording the association relationship between the sample identifier of the sample data and the annotation result in the annotation file in response to the labeling operation being performed on the displayed sample data, the annotation result characterizing the data category to which the sample data belongs, the annotation file including result statistical areas corresponding to the annotation results one by one, when recording the association relationship between the sample identifier and the annotation result, recording each sample identifier in the result statistical area of ​​the annotation result to be associated, and the sample identifier in each result statistical area is stored in a continuous storage area of ​​the storage medium; The screening module 153 is used to respond to receiving a screening instruction including a target annotation result, screen the sample data with the target annotation result in the sample data set according to the association between the sample identifier and the annotation result recorded in the annotation file, and process the screened sample data.

[0095] The sample data processing device in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0096] For the description of the features in the embodiment corresponding to the sample data processing device, reference may be made to the relevant description of the embodiment corresponding to the sample data processing method, which will not be described in detail here.

[0097] The embodiment of the present application also provides an electronic device, such as Fig.16As shown, it includes a memory 10 and a processor 20. The memory 10 stores a computer program. The processor 20 is configured to run the computer program to execute the steps in any of the above-mentioned sample data processing method embodiments.

[0098] An embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above-mentioned sample data processing method embodiments when running.

[0099] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

[0100] An embodiment of the present application further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps in any one of the above-mentioned sample data processing method embodiments are implemented.

[0101] An embodiment of the present application further provides another computer program product, including a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in any of the above-mentioned sample data processing method embodiments are implemented.

[0102] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may 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 this application.

[0103] The above is a detailed introduction to a sample data processing method, device, equipment, storage medium and program product provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core ideas of the present application. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the claims of the present application.

Claims

1. A sample data processing method, characterized in that: The method comprises: In response to receiving the first display instruction, display at least part of the sample data in the sample data set, the sample data set including a plurality of sample data, and the sample data having a sample identifier; In response to the displayed sample data being labeled, an association relationship between a sample identifier of the sample data and a labeling result is recorded in a labeling file, wherein the labeling result represents a data category to which the sample data belongs, and the labeling file includes a result statistical area corresponding to each of the labeling results. When recording the association relationship between the sample identifier and the labeling result, each sample identifier is recorded in a result statistical area of ​​the labeling result to be associated, and the sample identifier in each result statistical area is stored in a continuous storage area of ​​a storage medium; In response to receiving a screening instruction including a target annotation result, sample data having the target annotation result is screened in the sample data set according to the association between the sample identifier and the annotation result recorded in the annotation file, and the screened sample data is processed.

2. The method according to claim 1, characterized in that In the sample data set, the sample identification of the sample data is divided into a plurality of sample identification intervals, the annotation file further includes interval statistical areas corresponding one-to-one to the sample identification intervals, and each of the interval statistical areas respectively includes a result statistical area corresponding one-to-one to the annotation results; The recording of the association between the sample identifier of the sample data and the annotation result in the annotation file includes: Determine, according to the divided sample identification intervals, a target sample identification interval in which the sample identification is located; In the interval statistics area of ​​the target sample identification interval, the sample identification is recorded in the result statistics area of ​​the annotation result to be associated.

3. The method according to claim 2, characterized in that Before recording the association relationship between the sample identifier of the sample data and the annotation result in the annotation file, the method further includes: It is determined whether the sample data is the first labeled sample data in the sample data set. If so, the labeling file is created and the labeling file is set to a hidden state.

4. The method according to claim 3, characterized in that The determining whether the sample data is the first labeled sample data in the sample data set includes: Determine whether the annotation file exists, and if not, determine that the sample data is the first annotated sample data in the sample data set; or It is determined whether the value of the labeling state parameter corresponding to the sample data set is the first value. If so, it is determined that the sample data is the first labeled sample data in the sample data set, and the value of the labeling state parameter is changed to the second value.

5. The method according to claim 2, characterized in that: Before recording the association relationship between the sample identifier of the sample data and the annotation result in the annotation file, the method further includes: Check whether the interval statistical area of ​​the target sample identification interval is divided in the annotation file. If not, divide the interval statistical area of ​​the target sample identification interval in the annotation file, and divide the result statistical area corresponding to the annotation result one by one in the interval statistical area of ​​the target sample identification interval.

6. The method according to claim 1 or 2, characterized in that: When recording the association relationship between the sample identifier of the sample data and the annotation result in the annotation file, the method further includes: If there is a target sample identifier that is the same as the sample identifier of the sample data in the annotation file, and the annotation result associated with the target sample identifier is different from the annotation result of the sample data, the target sample identifier is deleted from the annotation file, and the association relationship between the sample identifier of the sample data and the annotation result is recorded in the annotation file.

7. The method according to claim 6, characterized in that The method further comprises: If the target sample identifier exists in the annotation file, and the annotation result associated with the target sample identifier is the same as the annotation result of the sample data, the target sample identifier continues to be retained in the annotation file, and the association relationship between the sample identifier and the annotation result of the sample data is stopped from being recorded in the annotation file.

8. The method according to claim 1, characterized in that: The method further comprises: In response to receiving the first exit instruction, stopping displaying the sample data; In response to receiving a second display instruction, at least part of the sample data in the sample data set is displayed again, wherein, if the sample data in the sample data set is updated after the display of the sample data is stopped, when at least part of the sample data in the sample data set is displayed again, an update prompt of the sample data is simultaneously displayed.

9. The method according to claim 8, characterized in that Determine whether the sample data in the sample data set is updated based on the following method: When the display of the sample data stops, recording the first data modification time of the sample data set in the annotation file, where the first data modification time is the last modification time of the sample data set when the display of the sample data stops; When displaying at least part of the sample data in the sample data set again, obtaining a second data modification time of the sample data set, where the second data modification time is the last modification time of the sample data set when the second display instruction is received; If the first data modification time and the second data modification time are different, it means that after the display of the sample data is stopped, the sample data in the sample data set is updated.

10. The method according to claim 1, characterized in that The method further comprises: In response to receiving a second exit instruction, stopping displaying the sample data; After the display of the sample data is stopped, if the sample data in the sample data set is updated, the changed target sample data is recorded; In response to receiving the third display instruction, at least part of the sample data in the sample data set is displayed again, and at the same time, an update prompt of the target sample data is displayed.

11. The method according to claim 10, characterized in that The simultaneously displaying the update prompt of the target sample data includes: Determining whether the target sample data has a labeling result according to the sample identifier of the target sample data, the association relationship between the sample identifier recorded in the labeling file and the labeling result; If the target sample data has a marked result, an update prompt of the target sample data is displayed at the same time; If the target sample data does not have a marking result, the update prompt of the target sample data will not be displayed.

12. The method according to claim 1, characterized in that After labeling the sample data in the sample data set, the method further includes: In response to receiving the statistical instruction, based on the association relationship between the sample identifiers and the annotation results recorded in the annotation file, counting the number of sample identifiers corresponding to each annotation result; Display the number of sample identifiers corresponding to each annotation result.

13. The method according to claim 1, characterized in that The sample data with the target annotation results are also used for page display in the sample data processing interface; The step of screening the sample data having the target annotation result in the sample data set includes: According to the number of sample data to be displayed on the current page, a corresponding number of sample data having the target marking results are screened from the sample data set.

14. The method according to claim 1, characterized in that The method is applied to a sample data processing system, which includes an initial data management module and a data set management module, wherein the initial data management module is used to store sample data, and the data set management module is used to process sample data; Before presenting at least part of the sample data in the sample data set, the method further includes: In response to one or more sample data in the initial data management module being selected, the selected sample data is imported into the data set management module as sample data in the sample data set, and the data set management module is used to display at least part of the sample data in the sample data set.

15. The method according to claim 14, characterized in that The step of importing the selected sample data into the data set management module includes: The storage path of the selected sample data in the initial data management module is imported into the data set management module. When displaying at least part of the sample data in the sample data set, the data set management module obtains the sample data to be displayed from the initial data management module based on the storage path.

16. The method according to claim 15, characterized in that After importing the storage path into the data set management module, the method further includes: In the initial data management module, the selected sample data is identified to distinguish the selected sample data from the unselected sample data in the initial data management module.

17. A sample data processing device, characterized in that: The device comprises: A display module, configured to display at least part of the sample data in the sample data set in response to receiving a first display instruction, wherein the sample data set includes a plurality of sample data and the sample data has a sample identifier; an association module, configured to record, in response to a labeling operation being performed on the displayed sample data, an association relationship between a sample identifier of the sample data and a labeling result in a labeling file, wherein the labeling result represents a data category to which the sample data belongs, and wherein the labeling file includes result statistical areas corresponding to the labeling results one by one, and when recording the association relationship between the sample identifier and the labeling result, each sample identifier is recorded in a result statistical area of ​​the labeling result to be associated, and the sample identifier in each result statistical area is stored in a continuous storage area of ​​a storage medium; A screening module is used to respond to receiving a screening instruction including a target annotation result, screen the sample data having the target annotation result in the sample data set according to the association relationship between the sample identifier and the annotation result recorded in the annotation file, and process the screened sample data.

18. An electronic device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the sample data processing method according to any one of claims 1 to 16 by executing the computer instructions.

19. A non-volatile computer-readable storage medium, characterized in that: The non-volatile computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the sample data processing method according to any one of claims 1 to 16.

20. A computer program product, characterized in that The method comprises computer instructions, wherein the computer instructions are used to cause a computer to execute the sample data processing method according to any one of claims 1 to 16.

Images