Big data information key domain data information extraction method

Abstract

The invention discloses a big data information key domain data information extraction method, and belongs to the technical field of data information processing. The technical problems that in an existing scheme, the extraction processing effect of new employee training big data information is poor, and the analysis evaluation effect is poor are solved. Processing and analyzing the overall training states of the new employees corresponding to different posts according to all the training data to obtain a plurality of target posts and a plurality of target training data; carrying out data processing and extraction of a non-entry reason dimension on all target posts obtained by processing, and carrying out data calculation and analysis mark combination on extracted data corresponding to different screened and classified non-entry reasons to obtain a first abnormal influence identification extraction set; and performing data processing and extraction of a failure data item dimension on all the target training data obtained by processing to obtain a second abnormal influence identification extraction set.

Description

Technical Field

[0001] The present invention relates to the technical field of data information processing, and in particular to a method for extracting key domain data information from big data. Background Art

[0002] The key domains of big data information refer to areas or data sets that are crucial in big data analysis and applications. These key domains usually contain data that have a significant impact on business decisions, market analysis, user behavior understanding, etc.

[0003] The implementation steps of data extraction from key domains of big data information can be divided into several main stages, from demand analysis to data application. These steps ensure the effective collection, processing and utilization of data. When the existing data extraction solutions for key domains of big data information are implemented, there is a single supervision and processing method and a single analysis and evaluation method for the extraction of information on new employee training and work, which leads to poor extraction and processing effects of big data information for new employee training and poor analysis and evaluation effects. Summary of the invention

[0004] The purpose of the present invention is to provide a method for extracting key domain data information of big data information, which is used to solve the technical problems of poor extraction and processing effect and poor analysis and evaluation effect of new employee training big data information in existing solutions.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] A method for extracting key domain data information of big data information, comprising:

[0007] Obtain the training data corresponding to all new employees in different positions, and process and analyze the overall training status of new employees corresponding to different positions based on all the training data to obtain several target positions and several target training data;

[0008] Data processing and extraction of the non-employment reason dimension are performed on all the target positions obtained, and the first abnormal impact identification extraction set of the corresponding dimension is obtained;

[0009] Performing data processing and extraction of failed data item dimensions on all target training data obtained through processing to obtain a second abnormal impact identification extraction set of corresponding dimensions;

[0010] The first abnormal impact identification extraction set and the second abnormal impact identification extraction set obtained by processing are pushed to the supervisors respectively, and targeted training abnormal extraction analysis prompts are implemented.

[0011] Preferably, when implementing different training supervision processes for all training data corresponding to different positions, the name, gender, education background, training start time, training end time and training results of the corresponding trainees in all training data for different positions are obtained in sequence;

[0012] When processing and extracting the training result dimension for the training data corresponding to all new employees in different positions, the training results in different training data of different positions are obtained in sequence and analyzed traversally;

[0013] If the training result is normal employment, a local training success instruction is generated and the total number of successful training CPi of the corresponding position is increased by one; i is a different position, i = 1, 2, 3, ..., n; n is a positive integer;

[0014] If the training result is failure to join the job, a local training failure instruction will be generated and the total number of training failures SPi for the corresponding position will be increased by one.

[0015] Preferably, when processing and analyzing the overall training status of new employees corresponding to different positions, the total number of successful training CPi and the total number of failed training SPi corresponding to different positions are calculated through the training status identification function, and the overall training identification ZPi corresponding to different positions is output;

[0016] Among them, the expression of the training state recognition function is In the formula, PB is the standard value of overall training requirements corresponding to different positions;

[0017] The overall training flag contains a value of 0 or 1, indicating whether the overall training status of new employees in the corresponding position is normal or abnormal;

[0018] According to the overall training flag with a value of 1, the corresponding position is marked as the target position, and the training data belonging to the training results of all target positions that have not yet entered the position are marked as target training data.

[0019] Preferably, the reasons for non-employment corresponding to the training results of non-employment in all target positions are obtained, and all the reasons for non-employment are classified and marked;

[0020] If the reason for non-joining is due to personal reasons of the employee, the non-joining employee will be marked as the first non-joining employee;

[0021] If the reason for non-employment is not due to personal reasons of the employee, the non-employment reason will be marked as the second non-employed employee;

[0022] All non-employment reasons corresponding to all second non-employed employees are obtained, and all non-employment reasons are sorted and combined, and all the same non-employment reasons in the sorted combination are deduplicated, and the total number of occurrences of different non-employment reasons is statistically updated to obtain a non-employment reasons statistical table.

[0023] Preferably, when processing the data on the influence of abnormal reasons corresponding to different reasons for non-employment according to the statistical table of reasons for non-employment, the total number of occurrences of different reasons for non-employment is calculated in turn by the formula Calculate and obtain the corresponding abnormal cause influence degree DYj; where NYj is the total number of occurrences of different reasons for non-employment; j is a different reason for non-employment, j=1, 2, 3, ..., m; m is a positive integer; B is the standard value of the abnormal cause influence.

[0024] Preferably, when data analysis is performed on the impact of abnormal reasons to determine the impact of abnormal reasons corresponding to different reasons for non-employment;

[0025] If the impact of the abnormal reason is 0, a general impact instruction of the abnormal reason is generated, and the corresponding non-employment reason is marked as a general non-employment reason;

[0026] If the impact of the abnormal reason is not 0, a special impact instruction of the abnormal reason is generated, and the corresponding non-employment reason is marked as a special non-employment reason;

[0027] All the special reasons for non-employment marked by the analysis are sorted and combined to obtain the first abnormal impact identification extraction set.

[0028] Preferably, data statistics are performed on different data items in all target training data to obtain corresponding failed gender statistical sequences, failed education statistical sequences, and failed training duration statistical sequences;

[0029] When performing abnormal identification and extraction of different elements in different statistical sequences, different elements in different statistical sequences are sequentially extracted through the formula Calculate and obtain the element influence ratio YLk of the corresponding element; where YZk is the value of different elements in different statistical sequences; k is the different elements in different statistical sequences, k=1, 2, 3, ..., p; p is a positive integer, which is the total number of elements in the statistical sequence.

[0030] Preferably, the difference in the element influence ratio between different elements in the statistical sequence is calculated and marked as the element influence difference YC. When determining the influence state of the abnormal data items between different elements according to the element influence difference, the influence differences of different elements in the statistical sequence are calculated by the element influence identification function and the corresponding abnormal data item influence degree YX is output;

[0031] Among them, the expression of the element influence identification function is Where, YC0 is the standard element influence difference;

[0032] The impact degree of the abnormal data item includes a value of 0 or 1, indicating whether the impact state of the abnormal data item between corresponding elements is normal or abnormal.

[0033] Preferably, according to the abnormal data item influence degree of 1, the maximum value element in the two calculation elements is marked as a high abnormal impact data item;

[0034] All high abnormal impact data items are sorted and combined to obtain a second abnormal impact identification and extraction set.

[0035] Preferably, the failure gender statistical sequence includes the total number of training failures for female students and the total number of training failures for male students;

[0036] The statistical series of failed academic qualifications includes the total number of failed college training, the total number of failed undergraduate training, and the total number of failed master's training;

[0037] The failed training duration statistical sequence includes the total number of failures during the first duration, the total number of failures during the second duration, and the total number of failures during the third duration.

[0038] Compared with the existing solutions, the present invention achieves the following beneficial effects:

[0039] The present invention processes and analyzes the overall training status of new employees corresponding to different positions based on all training data, and obtains several target positions and several target training data. It can not only obtain the overall training status of new employees corresponding to different positions, but also provide reliable supervision and analysis data support for subsequent data processing and extraction in different dimensions.

[0040] The present invention processes and extracts data from the dimension of reasons for non-employment for all target positions obtained through processing, and performs data calculation and analysis marking on the extracted data corresponding to different reasons for non-employment that are screened and classified, thereby obtaining different conventional reasons for non-employment and special reasons for non-employment, and extracts and combines all special reasons for non-employment to obtain a first abnormal impact identification extraction set, thereby improving the diversity and reliability of data extraction processing from the dimension of reasons for non-employment, and simultaneously providing reliable data support for extraction from the dimension of reasons for non-employment for subsequent training abnormal extraction and analysis prompts.

[0041] The present invention implements data processing and extraction of failed data item dimensions on all target training data obtained through processing, and performs data calculation on different element influence differences in different statistical sequences in turn through element influence recognition functions and outputs corresponding abnormal data item influence degrees, analyzes and screens abnormal data item influence states between different abnormal data items through the abnormal data item influence degrees, obtains high abnormal influence data items, sorts and combines all high abnormal influence data items, obtains a second abnormal influence recognition extraction set, thereby improving the diversity and reliability of data extraction processing of failed data item dimensions, and can also provide reliable failed data item dimension extraction data support for subsequent training abnormal extraction analysis prompts. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The present invention will be further described below in conjunction with the accompanying drawings.

[0043] Figure 1 The present invention is a flowchart of a method for extracting key domain data information of big data information.

[0044] Figure 2 This is a flowchart of the present invention for extracting the training result dimension from the training data corresponding to all new employees in different positions. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] like Figure 1 As shown, the present invention is a method for extracting key domain data information of big data information, comprising:

[0047] Obtain the training data corresponding to all new employees in different positions, and process and analyze the overall training status of new employees corresponding to different positions based on all training data to obtain several target positions and several target training data; including:

[0048] When implementing different training supervision methods for all training data corresponding to different positions, obtain the name, gender, education background, training start time, training end time and training results of the corresponding trainees in all training data of different positions in turn; the units corresponding to the start time and training end time are both days;

[0049] Among them, different positions include but are not limited to technical positions, sales positions and customer service positions;

[0050] like Figure 2 As shown, when the training result dimension is processed and extracted for the training data corresponding to all new employees in different positions, the training results in different training data of different positions are obtained in turn and traversed and analyzed;

[0051] Among them, the training results are normal employment or non-employment;

[0052] If the training result is normal employment, a local training success instruction is generated and the total number of successful training CPi of the corresponding position is increased by one; i is a different position, i = 1, 2, 3, ..., n; n is a positive integer;

[0053] If the training result is failure to join the company, a local training failure instruction is generated and the total number of training failures SPi for the corresponding position is increased by one;

[0054] When processing and analyzing the overall training status of new employees corresponding to different positions, the total number of successful training CPi and the total number of failed training SPi corresponding to different positions are calculated through the training status identification function, and the overall training identification ZPi corresponding to different positions is output;

[0055] Among them, the expression of the training state recognition function is Where PB is the standard value of the overall training requirements corresponding to different positions, which can be determined by the actual application requirements of the actual application scenario, or by the median value of all the overall training identifiers corresponding to different positions in history;

[0056] It should be noted that the overall training identifier is used to perform data calculation on all training result data corresponding to different positions, so as to digitally represent the overall training status of new employees corresponding to different positions;

[0057] The overall training flag contains a value of 0 or 1, indicating whether the overall training status of new employees in the corresponding position is normal or abnormal;

[0058] According to the overall training flag with a value of 1, the corresponding position is marked as the target position, and the training data belonging to the training results of all target positions that have not yet entered the position are marked as the target training data;

[0059] In an embodiment of the present invention, the overall training status of new employees corresponding to different positions is processed and analyzed based on all training data to obtain several target positions and several target training data. This can not only obtain the overall training status of new employees corresponding to different positions, but also provide reliable supervision and analysis data support for subsequent data processing and extraction in different dimensions.

[0060] Data processing and extraction of the non-employment reason dimension are performed on all the target positions obtained, and the first abnormal impact identification extraction set of the corresponding dimension is obtained; including:

[0061] Obtain the non-employment reasons corresponding to the training results of all target positions that have not been employed, and classify and mark all non-employment reasons;

[0062] Among them, the reasons for non-employment shall be determined by the trainers and training managers;

[0063] If the reason for non-joining is due to personal reasons of the employee, the non-joining employee will be marked as the first non-joining employee;

[0064] If the reason for non-employment is not due to personal reasons of the employee, the non-employment reason will be marked as the second non-employed employee;

[0065] In the embodiment of the present invention, by performing data analysis and classification marking on the reasons for non-employment, reliable screening data support can be provided for subsequent abnormal reasons impact analysis;

[0066] Obtain all non-joining reasons corresponding to all second non-joining employees, sort and combine all non-joining reasons, remove duplicates of all identical non-joining reasons in the sorted combination, and update the total number of occurrences of different non-joining reasons to obtain a non-joining reasons statistical table;

[0067] In the embodiment of the present invention, by sorting and combining all the reasons for non-employment marked by screening and removing duplicates, a statistical table of reasons for non-employment is obtained, thereby improving the reliability and diversity of the extraction and processing of reasons for non-employment;

[0068] When processing the data on the impact of abnormal reasons corresponding to different reasons for non-employment according to the statistical table of reasons for non-employment, the total number of occurrences of different reasons for non-employment is calculated in turn by the formula Calculate and obtain the corresponding abnormal cause influence degree DYj; where NYj is the total number of occurrences of different non-employment reasons; j is a different non-employment reason, j = 1, 2, 3, ..., m; m is a positive integer; B is the abnormal cause influence standard value; [*] is a rounding function, indicating obtaining the maximum integer that does not exceed the real number *;

[0069] It should be noted that the abnormal cause impact is used to perform data calculation on the extracted data corresponding to different reasons for non-employment, so as to digitally represent the abnormal cause impact corresponding to different reasons for non-employment;

[0070] When conducting data analysis on the impact of abnormal reasons to determine the impact of abnormal reasons corresponding to different reasons for non-employment;

[0071] If the impact of the abnormal reason is 0, a general impact instruction of the abnormal reason is generated, and the corresponding non-employment reason is marked as a general non-employment reason;

[0072] If the impact of the abnormal reason is not 0, a special impact instruction of the abnormal reason is generated, and the corresponding non-employment reason is marked as a special non-employment reason;

[0073] Sort and combine all the special reasons for non-employment marked by the analysis to obtain the first abnormal impact identification extraction set;

[0074] In the embodiment of the present invention, data processing and extraction of the dimension of reasons for non-employment are implemented for all target positions obtained through processing, and data calculation and analysis marking are performed on the extracted data corresponding to different reasons for non-employment that are screened and classified, so as to obtain different conventional reasons for non-employment and special reasons for non-employment, and then all special reasons for non-employment are extracted and combined to obtain the first abnormal impact identification extraction set, thereby improving the diversity and reliability of data extraction processing of the dimension of reasons for non-employment, and at the same time providing reliable data support for the extraction of the dimension of reasons for non-employment for subsequent training abnormal extraction analysis prompts.

[0075] Data processing and extraction of failed data item dimensions are performed on all target training data obtained through processing to obtain a second abnormal impact identification extraction set of corresponding dimensions; including:

[0076] Perform data statistics on different data items in all target training data to obtain corresponding statistical series of failed gender, failed education level, and failed training duration;

[0077] Among them, the failure gender statistical series includes the total number of female training failures and the total number of male training failures;

[0078] The statistical series of failed academic qualifications includes the total number of failed college training, the total number of failed undergraduate training, and the total number of failed master's training;

[0079] The statistical sequence of the duration of failed training includes the total number of failures during the first duration, the total number of failures during the second duration, and the total number of failures during the third duration;

[0080] Specifically, the first duration may be 1-7 days, the second duration may be 8-14 days, and the third duration may be more than 14 days;

[0081] In the embodiment of the present invention, by processing and counting the failed gender statistical sequence, the failed education statistical sequence, and the failed training duration statistical sequence for all target training data, the reliability and comprehensiveness of the extraction and processing of different data items can be effectively improved;

[0082] When performing abnormal identification and extraction of different elements in different statistical sequences, different elements in different statistical sequences are sequentially extracted through the formula Calculate and obtain the element influence ratio YLk of the corresponding element; where YZk is the value of different elements in different statistical sequences; k is different elements in different statistical sequences, k = 1, 2, 3, ..., p; p is a positive integer, which is the total number of elements in the corresponding statistical sequence;

[0083] And, the difference in the element influence ratio between different elements in the statistical sequence is calculated and marked as the element influence difference YC. When determining the influence state of abnormal data items between different elements according to the element influence difference, the influence difference of different elements in the statistical sequence is calculated by the element influence identification function and the corresponding abnormal data item influence degree YX is output;

[0084] Among them, the expression of the element influence identification function is Where YC0 is the standard element influence difference, which can be determined by the actual application requirements of the actual application scenario, or by the median of the influence of all abnormal data items in the history corresponding to different abnormal data items;

[0085] It should be noted that the abnormal data item influence degree is used to perform data calculation and digital representation on the abnormal data item influence status between different abnormal data items;

[0086] The impact degree of abnormal data items contains a value of 0 or 1, indicating whether the abnormal data items between corresponding elements have a normal or abnormal impact state;

[0087] According to the influence degree of the abnormal data item with a value of 1, the maximum value element in the two calculation elements is marked as a high abnormal influence data item;

[0088] For example, the total number of failed female training and the total number of failed male training contained in the failed gender statistical sequence have corresponding element influence proportions of 47% and 53% respectively, and the element influence difference between the two elements is 6%; at this time, the element influence difference of 6% is not greater than the standard element influence difference of 20%, so it is determined that the influence status of the abnormal data item between the total number of failed female training and the total number of failed male training contained in the failed gender statistical sequence is normal;

[0089] If the corresponding element influence ratios are 37% and 63% respectively, the element influence difference between the two elements is 26%; at this time, the element influence difference of 26% is greater than the standard element influence difference of 20%, and the abnormal data item influence status between the total number of female training failures and the total number of male training failures contained in the failed gender statistical sequence is determined to be abnormal, and at this time, the total number of male training failures with an element influence ratio of 63% is a high abnormal influence data item;

[0090] Sort and combine all high abnormal impact data items to obtain a second abnormal impact identification and extraction set;

[0091] In an embodiment of the present invention, data processing and extraction of failed data item dimensions are performed on all target training data obtained through processing, and data calculations are performed in sequence on different element influence differences in different statistical sequences through element influence recognition functions, and the corresponding abnormal data item influence degrees are output. The abnormal data item influence states between different abnormal data items are analyzed and screened through the abnormal data item influence degrees to obtain high abnormal influence data items, and all high abnormal influence data items are sorted and combined to obtain a second abnormal influence identification and extraction set, thereby improving the diversity and reliability of data extraction processing of failed data item dimensions, and also providing reliable failed data item dimension extraction data support for subsequent training abnormal extraction analysis prompts.

[0092] The first abnormal impact identification extraction set and the second abnormal impact identification extraction set obtained by processing are pushed to the supervisors respectively, and targeted training abnormal extraction analysis prompts are implemented.

[0093] Among them, targeted training anomaly extraction analysis prompts are implemented based on all special reasons for non-employment in the first abnormal impact identification and extraction set and all high abnormal impact data items in the second abnormal impact identification and extraction set, which can provide reliable multi-dimensional training extraction data support for the implementation and management of subsequent new employee training.

[0094] In addition, the formulas involved in the above are all dimensionless and numerical calculations. They are a formula that is closest to the actual situation obtained by collecting a large amount of data and simulating it with simulation software.

[0095] In the several embodiments provided by the present invention, it should be understood that the disclosed method can be implemented in other ways. For example, the above-described embodiments of the invention are only illustrative, for example, the division of modules is only a logical function division, and there may be other division methods in actual implementation.

[0096] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0097] In addition, each functional module in each embodiment of the present invention may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of hardware plus software functional modules.

[0098] It is obvious to a person skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the essential characteristics of the present invention.

[0099] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for extracting key domain data information from big data, characterized in that: include: Obtain the training data corresponding to all new employees in different positions, and process and analyze the overall training status of new employees corresponding to different positions based on all the training data to obtain several target positions and several target training data; Data processing and extraction of the non-employment reason dimension are performed on all the target positions obtained, and the first abnormal impact identification extraction set of the corresponding dimension is obtained; Performing data processing and extraction of failed data item dimensions on all target training data obtained through processing to obtain a second abnormal impact identification extraction set of corresponding dimensions; The first abnormal impact identification extraction set and the second abnormal impact identification extraction set obtained by processing are pushed to the supervisors respectively, and targeted training abnormal extraction analysis prompts are implemented.

2. The method for extracting key domain data information from big data information according to claim 1, characterized in that: When implementing different forms of training supervision processing for all training data corresponding to different positions, the name, gender, education background, training start time, training end time and training results of the corresponding trainees in all training data for different positions are obtained in sequence; When processing and extracting the training result dimension for the training data corresponding to all new employees in different positions, the training results in different training data of different positions are obtained in sequence and analyzed traversally; If the training result is normal employment, a local training success instruction is generated and the total number of successful training CPi of the corresponding position is increased by one; i is a different position, i = 1, 2, 3, ..., n; n is a positive integer; If the training result is failure to join the job, a local training failure instruction will be generated and the total number of training failures SPi for the corresponding position will be increased by one.

3. The method for extracting key domain data information from big data information according to claim 2 is characterized in that: When processing and analyzing the overall training status of new employees corresponding to different positions, the total number of successful training CPi and the total number of failed training SPi corresponding to different positions are calculated through the training status identification function, and the overall training identification ZPi corresponding to different positions is output; Among them, the expression of the training state recognition function is In the formula, PB is the standard value of overall training requirements corresponding to different positions; The overall training flag contains a value of 0 or 1, indicating whether the overall training status of new employees in the corresponding position is normal or abnormal; According to the overall training flag with a value of 1, the corresponding position is marked as the target position, and the training data belonging to the training results of all target positions that have not yet entered the position are marked as target training data.

4. The method for extracting key domain data information from big data information according to claim 3 is characterized in that: Obtain the non-employment reasons corresponding to the training results of all target positions that have not been employed, and classify and mark all non-employment reasons; If the reason for non-joining is due to personal reasons of the employee, the non-joining employee will be marked as the first non-joining employee; If the reason for non-employment is not due to personal reasons of the employee, the non-employment reason will be marked as the second non-employed employee; All non-employment reasons corresponding to all second non-employed employees are obtained, and all non-employment reasons are sorted and combined, and all the same non-employment reasons in the sorted combination are deduplicated, and the total number of occurrences of different non-employment reasons is statistically updated to obtain a non-employment reasons statistical table.

5. The method for extracting key domain data information from big data information according to claim 4 is characterized in that: When processing the data on the impact of abnormal reasons corresponding to different reasons for non-employment according to the statistical table of reasons for non-employment, the total number of occurrences of different reasons for non-employment is calculated in turn by the formula Calculate and obtain the corresponding abnormal cause influence degree DYj; where NYj is the total number of occurrences of different reasons for non-employment; j is a different reason for non-employment, j=1, 2, 3, ..., m; m is a positive integer; B is the standard value of the abnormal cause influence.

6. The method for extracting key domain data information from big data information according to claim 5, characterized in that: When conducting data analysis on the impact of abnormal reasons to determine the impact of abnormal reasons corresponding to different reasons for non-employment; If the impact of the abnormal reason is 0, a general impact instruction of the abnormal reason is generated, and the corresponding non-employment reason is marked as a general non-employment reason; If the impact of the abnormal reason is not 0, a special impact instruction of the abnormal reason is generated, and the corresponding non-employment reason is marked as a special non-employment reason; All the special reasons for non-employment marked by the analysis are sorted and combined to obtain the first abnormal impact identification extraction set.

7. The method for extracting key domain data information from big data information according to claim 6, characterized in that: Perform data statistics on different data items in all target training data to obtain corresponding statistical series of failed gender, failed education level, and failed training duration; When performing abnormal identification and extraction of different elements in different statistical sequences, different elements in different statistical sequences are sequentially extracted through the formula Calculate and obtain the element influence ratio YLk of the corresponding element; where YZk is the value of different elements in different statistical sequences; k is the different elements in different statistical sequences, k=1, 2, 3, ..., p; p is a positive integer, which is the total number of elements in the statistical sequence.

8. The method for extracting key domain data information from big data information according to claim 7, characterized in that: Calculate the difference in the element influence ratio between different elements in the statistical sequence and mark it as the element influence difference YC. When determining the influence status of abnormal data items between different elements according to the element influence difference, calculate the influence difference of different elements in the statistical sequence through the element influence identification function and output the corresponding abnormal data item influence degree YX; Among them, the expression of the element influence identification function is Where, YC0 is the standard element influence difference; The impact degree of the abnormal data item includes a value of 0 or 1, indicating whether the impact state of the abnormal data item between corresponding elements is normal or abnormal.

9. The method for extracting key domain data information from big data information according to claim 8, characterized in that: According to the influence degree of the abnormal data item with a value of 1, the maximum value element in the two calculation elements is marked as a high abnormal influence data item; All high abnormal impact data items are sorted and combined to obtain a second abnormal impact identification and extraction set.

10. The method for extracting key domain data information from big data information according to claim 7, characterized in that: The failure gender statistics series includes the total number of female training failures and the total number of male training failures; The statistical series of failed academic qualifications includes the total number of failed college training, the total number of failed undergraduate training, and the total number of failed master's training; The failed training duration statistical sequence includes the total number of failures during the first duration, the total number of failures during the second duration, and the total number of failures during the third duration.

Images