Electronic information data classification method, system and medium for archive management

Abstract

The present invention relates to the technical field of data clustering, and in particular to an electronic information data classification method, system and medium for archive management, the method comprising: obtaining electronic information data of each personnel archive to be classified, preprocessing the electronic information data to obtain word vectors corresponding to each word segmentation in the electronic information data; inputting the information of the current task and the electronic information data of the personnel archive to be classified into a trained classification network, outputting the task demand index of each skill of the personnel corresponding to the personnel archive to be classified under the condition of the current task; obtaining the skill mastering level and learning ability of each skill according to the task demand index, word vector, electronic information data and the skills required for the current task; obtaining the clustering distance according to the skills required for the current task, the skill mastering level and the learning ability; performing clustering processing according to the clustering distance; and saving the clustering result. The present invention can improve the accuracy of the classification result of the archive.

Description

Technical Field

[0001] The present invention relates to the technical field of data clustering, and in particular to an electronic information data classification method, system and medium for archive management. Background Art

[0002] With the rapid development of information technology, the transformation from paper archives to digital archives has become an inevitable trend. Digitization not only solves the limitation of physical storage space, but also improves the accessibility and long-term preservation of personnel archives. Traditional archive management focuses on the physical management of documents and file-level organization, while the era of big data requires archivists to have data thinking, that is, to consider problems from the perspective of full data mode, inclusive heterogeneity and focus on relevant relationships.

[0003] Personnel files correspond to personnel management data classification methods. It is a way to organize and manage data in electronic files, records and information systems according to certain standards and rules so that they can be stored, retrieved and used efficiently. Due to the huge amount and variety of electronic information data, a reasonable classification method is essential.

[0004] However, existing classification methods usually perform clustering directly based on the characteristics of electronic information data. When faced with archive classification requirements under different task objectives, classification based solely on data characteristics is often difficult to achieve ideal classification results, making the archive classification results inaccurate. Summary of the invention

[0005] In order to solve the technical problem that the classification results of archives are not accurate enough, the purpose of the present invention is to provide an electronic information data classification method, system and medium for archive management. The technical solution adopted is as follows:

[0006] In a first aspect of the present disclosure, there is provided an electronic information data classification method for archive management, the method comprising:

[0007] Obtaining electronic information data of each personnel file to be classified, and preprocessing the electronic information data to obtain word vectors corresponding to each word segment in the electronic information data;

[0008] Inputting the information of the current task and the electronic information data of the personnel file to be classified into the trained classification network, and outputting the task demand index of each skill of the personnel corresponding to the personnel file to be classified under the condition of the current task;

[0009] According to the task requirement index of each skill of the personnel corresponding to the personnel files to be classified, the word vector corresponding to each word segmentation, the electronic information data and the skills required for the current task, the skill mastery level of each person corresponding to the personnel files to be classified for each skill required for the current task and the learning ability of each person corresponding to the personnel files to be classified under the condition of the current task are obtained;

[0010] According to the skills required for the current task, the skill mastering level and the learning ability, obtaining the clustering distance of the electronic information data of every two personnel files to be classified;

[0011] Clustering the electronic information data of each of the personnel files to be classified according to the clustering distance;

[0012] The clustering result is saved.

[0013] In one embodiment, the preprocessing of the electronic information data to obtain a word vector corresponding to each word segment in the electronic information data includes:

[0014] Performing word segmentation processing on the text in the electronic information data to obtain word segments corresponding to the text;

[0015] Removing stop words from each of the participles;

[0016] Performing Chinese word form restoration processing on each of the word segments after removing the stop words;

[0017] Each of the word segments after Chinese lemma restoration is converted into a word vector.

[0018] In one embodiment, before inputting the information of the current task and the electronic information data of the personnel files to be classified into the trained neural network, the method further includes:

[0019] Acquire a sample set, the sample set including: information of the current task, sample skill keywords of the field to which the preset personnel files to be classified correspond to the personnel files, and preset task requirement indexes of the marked sample skill keywords;

[0020] Dividing the sample set into a training set and a validation set;

[0021] Inputting the training set into the original classification network, iteratively training the original classification network using a gradient descent method, and updating the parameters in the original classification network using a preset loss function and the verification set;

[0022] When the preset loss function converges, the iterative training of the original classification network is stopped to obtain the trained classification network.

[0023] In one embodiment, the step of obtaining the skill mastery level of each skill required for the current task of each person corresponding to the personnel file to be classified according to the task requirement index of each skill of the person corresponding to the personnel file to be classified, the word vector corresponding to each word segmentation, the electronic information data and the skills required for the current task includes:

[0024] The following steps are performed for each electronic information data of the personnel file to be classified and each skill required for the current task:

[0025] According to the word vector corresponding to each text segmentation and the current electronic information data, the cosine similarity between each text segmentation and the current skill keyword in the personal skill module in the current electronic information data is calculated; the current electronic information data is the electronic information data of the personnel file to be classified, and the text segmentation represents the segmentation in the current electronic information data; the current skill keyword is the keyword corresponding to the current skill required for the current task;

[0026] According to the personal skill module and the target segmentation in the current electronic information data, the mastery position coefficient of the current skill of the person corresponding to the current personnel file to be classified is obtained; the target segmentation is the text segmentation corresponding to the largest cosine similarity;

[0027] According to the work experience module in the current electronic information data and the target segmentation, obtaining the skill mastery degree and skill necessity degree of the current skill in each work experience in the work experience module;

[0028] According to the skill mastery degree and the skill necessity degree, obtaining the objective skill mastery level of the current skill of the person corresponding to the personnel file to be classified;

[0029] According to the mastering position coefficient and the objective mastering level of the skill, the skill mastering level of the current skill of the person corresponding to the personnel file to be classified is obtained.

[0030] In one embodiment, the step of obtaining the mastery position coefficient of the current skill of the person corresponding to the current personnel file to be classified according to the personal skill module in the current electronic information data and the target segmentation includes:

[0031] Obtaining a first number of all the text segmentations before the target segmentation in a target paragraph and a second number of all the segmentations included in the target paragraph; the target paragraph is the paragraph where the target segmentation in the personal skill module is located;

[0032] Obtaining the number of previous paragraphs before the target paragraph in the personal skill module and the total number of paragraphs in the personal skill module;

[0033] The mastery position coefficient of the current skill of the person corresponding to the personnel file to be classified is obtained according to the first number, the second number, the previous number of paragraphs and the total number of paragraphs.

[0034] In one embodiment, according to the work experience module in the current electronic information data and the target segmentation, obtaining the skill mastery level and skill necessity level of the current skill in each work experience in the work experience module includes:

[0035] Obtaining a third number of all the text segmentations included in each paragraph in the work experience module;

[0036] Calculate the information entropy between the word vectors of each of the text segmentations in each paragraph in the work experience module;

[0037] According to the third quantity and the information entropy, obtaining the skill mastery level reflected by each of the work experiences;

[0038] Obtain the target paragraph in which the target word appears in the work experience module;

[0039] Obtaining the target sentence where the target word is located in each of the target paragraphs;

[0040] Obtaining a fourth number of the text segmentations contained in each of the target sentences;

[0041] Obtaining a fifth number of the text segmentations corresponding to the other skill keywords contained in each of the target sentences;

[0042] The skill necessity degree of the current skill in each of the work experiences is obtained based on the fourth quantity and the fifth quantity.

[0043] In one embodiment, according to the task requirement index of each skill of the personnel corresponding to the personnel file to be classified, the word vector corresponding to each word segmentation, the electronic information data and the skills required for the current task, the learning ability of each person corresponding to the personnel file to be classified under the condition of the current task is obtained, including:

[0044] The following steps are performed for each electronic information data of the personnel file to be classified:

[0045] Acquire age data in the current electronic information data;

[0046] Acquire a sixth number of the text segmentations included in the education experience part of the current electronic information data;

[0047] acquiring a first learning ability according to the age data and the sixth quantity;

[0048] Obtain the working hours of each of the working experiences;

[0049] Calculate the learning ability reflected in each of the work experiences based on the working hours of each of the work experiences, the number of all the skills included in each of the work experiences, the necessary degree of all the skills, and the objective mastery level of all the skills;

[0050] Acquire the learning ability reflected by the current electronic information data according to the first learning ability and the learning abilities reflected by each of the work experiences;

[0051] Sorting the task requirement indexes corresponding to the skills in a preset order to obtain a first sequence;

[0052] Sorting the skill mastering level of each skill in the current electronic information data according to the preset order to obtain a second sequence;

[0053] Obtaining a correlation coefficient between the first sequence and the second sequence;

[0054] According to the learning ability reflected by the current electronic information data and the correlation coefficient, the learning ability of the personnel corresponding to the current personnel file to be classified under the condition of the current task is obtained.

[0055] In one embodiment, the step of obtaining the clustering distance of the electronic information data of every two personnel files to be classified according to the skills, the skill mastering level and the learning ability includes:

[0056] The following steps are performed for each electronic information data of two personnel files to be classified:

[0057] Obtain the same skills of the personnel corresponding to the two personnel files to be classified;

[0058] From the same skills of the personnel corresponding to the two personnel files to be classified, obtaining a target skill included in the skills required for the current task;

[0059] The clustering distance of the electronic information data of the two personnel files to be classified is obtained according to the target skills, the skill mastery levels of the personnel corresponding to the two personnel files to be classified for each target skill and the learning abilities of the personnel corresponding to the two personnel files to be classified.

[0060] In a second aspect of the present disclosure, there is provided an electronic information data classification system for archive management, the system comprising:

[0061] A first acquisition module is used to acquire the electronic information data of each personnel file to be classified, and pre-process the electronic information data to obtain the word vector corresponding to each word segment in the electronic information data;

[0062] The second acquisition module is used to input the information of the current task and the electronic information data of the personnel file to be classified into the trained classification network, and output the task demand index of each skill of the personnel corresponding to the personnel file to be classified under the condition of the current task;

[0063] The third acquisition module is used to obtain the skill mastery level of each person corresponding to the personnel file to be classified for each skill required for the current task and the learning ability of each person corresponding to the personnel file to be classified under the condition of the current task according to the task requirement index of each skill of the person corresponding to the personnel file to be classified, the word vector corresponding to each word segmentation, the electronic information data and the skills required for the current task;

[0064] A fourth acquisition module, for acquiring the clustering distance of the electronic information data of every two personnel files to be classified according to the skills required for the current task, the skill mastering level and the learning ability;

[0065] A clustering processing module, used for clustering the electronic information data of each of the personnel files to be classified according to the clustering distance;

[0066] A saving module is used to save the clustering result.

[0067] In one embodiment, in preprocessing the electronic information data to obtain word vectors corresponding to each word segment in the electronic information data, the first acquisition module is specifically used to:

[0068] Performing word segmentation processing on the text in the electronic information data to obtain word segments corresponding to the text;

[0069] Removing stop words from each of the participles;

[0070] Performing Chinese word form restoration processing on each of the word segments after removing the stop words;

[0071] Each of the word segments after Chinese lemma restoration is converted into a word vector.

[0072] In one embodiment, before inputting the information of the current task and the electronic information data of the personnel file to be classified into the trained neural network, the second acquisition module is further used to:

[0073] Acquire a sample set, the sample set including: information of the current task, sample skill keywords of the field to which the preset personnel files to be classified correspond to the personnel files, and preset task requirement indexes of the marked sample skill keywords;

[0074] Dividing the sample set into a training set and a validation set;

[0075] Inputting the training set into the original classification network, iteratively training the original classification network using a gradient descent method, and updating the parameters in the original classification network using a preset loss function and the verification set;

[0076] When the preset loss function converges, the iterative training of the original classification network is stopped to obtain the trained classification network.

[0077] In one embodiment, in the aspect of acquiring the skill mastery level of each skill required for the current task of each person corresponding to the personnel file to be classified according to the task requirement index of each skill of the person corresponding to the personnel file to be classified, the word vector corresponding to each word segmentation, the electronic information data and the skills required for the current task, the third acquisition module is specifically used to:

[0078] The following steps are performed for each electronic information data of the personnel file to be classified and each skill required for the current task:

[0079] According to the word vector corresponding to each text segmentation and the current electronic information data, the cosine similarity between each text segmentation and the current skill keyword in the personal skill module in the current electronic information data is calculated; the current electronic information data is the electronic information data of the personnel file to be classified, and the text segmentation represents the segmentation in the current electronic information data; the current skill keyword is the keyword corresponding to the current skill required for the current task;

[0080] According to the personal skill module and the target segmentation in the current electronic information data, the mastery position coefficient of the current skill of the person corresponding to the current personnel file to be classified is obtained; the target segmentation is the text segmentation corresponding to the largest cosine similarity;

[0081] According to the work experience module in the current electronic information data and the target segmentation, obtaining the skill mastery degree and skill necessity degree of the current skill in each work experience in the work experience module;

[0082] According to the skill mastery degree and the skill necessity degree, obtaining the objective skill mastery level of the current skill of the person corresponding to the personnel file to be classified;

[0083] According to the mastering position coefficient and the objective mastering level of the skill, the skill mastering level of the current skill of the person corresponding to the personnel file to be classified is obtained.

[0084] In one embodiment, in the aspect of obtaining the mastery position coefficient of the current skill of the person corresponding to the current personnel file to be classified according to the personal skill module in the current electronic information data and the target segmentation, the third acquisition module is specifically used to:

[0085] Obtaining a first number of all the text segmentations before the target segmentation in a target paragraph and a second number of all the segmentations included in the target paragraph; the target paragraph is the paragraph where the target segmentation in the personal skill module is located;

[0086] Obtaining the number of previous paragraphs before the target paragraph in the personal skill module and the total number of paragraphs in the personal skill module;

[0087] The mastery position coefficient of the current skill of the person corresponding to the personnel file to be classified is obtained according to the first number, the second number, the previous number of paragraphs and the total number of paragraphs.

[0088] In one embodiment, in terms of acquiring the skill mastery level and skill necessity level of the current skill in each work experience of the work experience module and the target segmentation in the current electronic information data, the third acquisition module is specifically used to:

[0089] Obtaining a third number of all the text segmentations included in each paragraph in the work experience module;

[0090] Calculate the information entropy between the word vectors of each of the text segmentations in each paragraph in the work experience module;

[0091] According to the third quantity and the information entropy, obtaining the skill mastery level reflected by each of the work experiences;

[0092] Obtain the target paragraph in which the target word appears in the work experience module;

[0093] Obtaining the target sentence where the target word is located in each of the target paragraphs;

[0094] Obtaining a fourth number of the text segmentations contained in each of the target sentences;

[0095] Obtaining a fifth number of the text segmentations corresponding to the other skill keywords contained in each of the target sentences;

[0096] The skill necessity degree of the current skill in each of the work experiences is obtained based on the fourth quantity and the fifth quantity.

[0097] In one embodiment, in terms of acquiring the learning ability of each person corresponding to the personnel files to be classified under the condition of the current task according to the task requirement index of each skill of the person corresponding to the personnel files to be classified, the word vector corresponding to each word segmentation, the electronic information data and the skills required for the current task, the third acquisition module is specifically used to:

[0098] The following steps are performed for each electronic information data of the personnel file to be classified:

[0099] Acquire age data in the current electronic information data;

[0100] Acquire a sixth number of the text segmentations included in the education experience part of the current electronic information data;

[0101] acquiring a first learning ability according to the age data and the sixth quantity;

[0102] Obtain the working hours of each of the working experiences;

[0103] Calculate the learning ability reflected in each of the work experiences based on the working hours of each of the work experiences, the number of all the skills included in each of the work experiences, the necessary degree of all the skills, and the objective mastery level of all the skills;

[0104] Acquire the learning ability reflected by the current electronic information data according to the first learning ability and the learning abilities reflected by each of the work experiences;

[0105] Sorting the task requirement indexes corresponding to the skills in a preset order to obtain a first sequence;

[0106] Sorting the skill mastering level of each skill in the current electronic information data according to the preset order to obtain a second sequence;

[0107] Obtaining a correlation coefficient between the first sequence and the second sequence;

[0108] According to the learning ability reflected by the current electronic information data and the correlation coefficient, the learning ability of the personnel corresponding to the current personnel file to be classified under the condition of the current task is obtained.

[0109] In one embodiment, in the aspect of obtaining the clustering distance of the electronic information data of every two personnel files to be classified according to the skills, the skill mastering level and the learning ability, the fourth obtaining module is specifically used for:

[0110] The following steps are performed for each electronic information data of two personnel files to be classified:

[0111] Obtain the same skills of the personnel corresponding to the two personnel files to be classified;

[0112] From the same skills of the personnel corresponding to the two personnel files to be classified, obtaining a target skill included in the skills required for the current task;

[0113] The clustering distance of the electronic information data of the two personnel files to be classified is obtained according to the target skills, the skill mastery levels of the personnel corresponding to the two personnel files to be classified for each target skill and the learning abilities of the personnel corresponding to the two personnel files to be classified.

[0114] In a third aspect of the present disclosure, there is provided a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the electronic information data classification method for archive management as described in any one of the embodiments of the first aspect.

[0115] The present invention has the following beneficial effects: the present application evaluates the importance of the current task to different skill requirements through a classification network; analyzes the skill mastery level of different personnel in the electronic information data of each personnel file to be classified, and analyzes the learning ability of the personnel corresponding to each personnel file to be classified in combination with the importance of the current task to each skill requirement, and further optimizes the clustering distance measurement method of the electronic information data of each personnel file to be classified under the current task requirement conditions; finally, clustering processing is performed on the electronic information data of each personnel file to be classified. Compared with the existing method, the clustering distance measurement process takes into account the skills required for the current task, the skill mastery level of each person corresponding to each personnel file to be classified for each skill required for the current task, and the learning ability of each person corresponding to each personnel file to be classified under the current task conditions. Since the clustering distance evaluation method between the electronic information data of each personnel file to be classified is optimized, the clustering result is more in line with the needs of the current task, thereby reducing the cost of human resource decision-making and improving planning efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0116] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0117] Figure 1 A method flow of an electronic information data classification method for archive management provided by an embodiment of the present invention Figure 1 ;

[0118] Figure 2 A method flow of an electronic information data classification method for archive management provided by an embodiment of the present invention Figure 2 ;

[0119] Figure 3 A method flow of an electronic information data classification method for archive management provided by an embodiment of the present invention Figure 3 ;

[0120] Figure 4 A method flow of an electronic information data classification method for archive management provided by an embodiment of the present invention Figure 4 ;

[0121] Figure 5 A method flow of an electronic information data classification method for archive management provided by an embodiment of the present invention Figure 5 ;

[0122] Figure 6 A method flow of an electronic information data classification method for archive management provided by an embodiment of the present invention Figure 6 ;

[0123] Figure 7 A structural block diagram of an electronic information data classification system for archive management provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0124] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the electronic information data classification method, system and medium for archive management proposed by the present invention, its specific implementation method, structure, features and effects in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0125] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0126] With the acceleration of the informatization process in various fields of society, especially the popularization and application of emerging technologies such as big data, cloud computing, and artificial intelligence, the storage, management, and use of electronic archives have changed: in order to increase the efficiency of personnel deployment and reduce the cost of human resource decision-making in the task allocation process, it is necessary to classify the archives based on the data in the personnel electronic archives. When faced with different task requirements, how to reasonably classify these data to ensure that they can be efficiently and accurately managed throughout the entire process of storage, retrieval, sharing, and archiving has become a major challenge in the process of informatization.

[0127] The following is a detailed description of a method, system and medium for electronic information data classification for archive management provided by the present invention in conjunction with the accompanying drawings.

[0128] See also Figure 1 , which shows a method flow chart of an electronic information data classification method for archive management provided by an embodiment of the present invention, such as Figure 1 As shown, the method includes the following steps S101-S106:

[0129] S101, obtaining electronic information data of each personnel file to be classified, and preprocessing the electronic information data to obtain word vectors corresponding to each word segment in the electronic information data.

[0130] Among them, the personnel files corresponding to the personnel usually contain different categories of content such as the employee's personal information, work history, professional skills, etc. This information is usually saved in the form of text files, such as Word documents, Markdown documents and other formats. The electronic information data of the personnel files represents the personnel files corresponding to the personnel saved in the form of text.

[0131] In one embodiment, Figure 2 As shown, preprocessing the electronic information data to obtain the word vector corresponding to each word in the electronic information data includes the following sub-steps S1011-S1014:

[0132] S1011, performing word segmentation processing on the text in the electronic information data to obtain word segments corresponding to the text.

[0133] In this step, the text is segmented. Commonly used segmentation tools include jieba segmentation, ictclas segmentation, etc.

[0134] S1012: Remove stop words from each word segment.

[0135] You can use the Internet to search for Chinese stop word lists to remove stop words from the text after word segmentation.

[0136] S1013, performing Chinese word form restoration processing on each word segment after removing stop words.

[0137] The part-of-speech tagging library pynlpir can be used to perform Chinese word form restoration on the extracted word segments.

[0138] S1014, converting each word segment after Chinese word form restoration processing into a word vector.

[0139] The word2vec method can be used to convert different word segmentations into word vector form.

[0140] S102, inputting the information of the current task and the electronic information data of the personnel files to be classified into the trained neural network, and outputting the task requirement index of each skill of the personnel corresponding to the personnel files to be classified under the condition of the current task.

[0141] The information of the current task includes the purpose of this file classification and the tasks that the personnel corresponding to the selected personnel files need to complete, which may include: task description, requirements, planning, etc.

[0142] In the process of classifying personnel according to personnel files, in order to improve the efficiency of human resource decision-making, it is necessary to optimize the existing classification process in combination with the skill requirements for personnel in different classification tasks. For the current classification task requirements, the degree of demand for various skills in the field of the personnel corresponding to the personnel files to be classified can be evaluated by training a neural network. In one embodiment, before executing step S102, the original neural network needs to be trained to obtain the trained neural network to be used in step S102. The training of the original neural network includes the following steps A1-A4:

[0143] A1. Obtain a sample set, which includes: information about the current task, sample skill keywords of the field to which the preset personnel files to be classified correspond to the personnel files, and preset task demand indexes of the marked sample skill keywords.

[0144] Get the information of the current task, and select sample skill keywords in the personnel profile field corresponding to the preset personnel profile to be classified. For example, for backend developers, their skill keywords can be divided into Java, Python, Mysql, SSM, SpringBoot, VUE, jquery, Linux, Redis, nginx, etc.

[0145] A set of task information in the same field is collected, and each skill is scored manually based on the above characteristics. The scoring values ​​are: 0.1, 0.2, 0.3, ... 1, a total of 10 score types, corresponding to 10 levels of task requirements. The scoring result is the task requirement index. The higher the index, the higher the degree of mastery of the skill requirements for this type of sample under the current task.

[0146] A2. Divide the sample set into training set and validation set.

[0147] A3. Input the training set into the original classification network, perform iterative training on the original classification network using the gradient descent method, and update the parameters in the original classification network using the preset loss function and the validation set;

[0148] A4. When the preset loss function converges, stop the iterative training of the original classification network to obtain the trained classification network.

[0149] The classification network in the present disclosure may adopt a 5-layer fully connected neural network structure. In other embodiments, it may be set to a neural network with other structures. This embodiment does not make specific limitations. The loss function may be a cross entropy function.

[0150] The labeled sample set is divided into a training set and a validation set in a ratio of 7:3. The training set data is input into the original classification network for training. The gradient descent method is used for training until the loss function converges and the training of the classification network is completed.

[0151] At the entrance of the trained classification network, the information of the current task is input into the trained neural network, and the task demand index A of each skill in the field to which the personnel file to be classified currently belongs is output.

[0152] S103. Obtain the skill mastery level of each skill required for the current task of each person corresponding to each personnel file to be classified and the learning ability of each person corresponding to each personnel file to be classified under the conditions of the current task based on the task requirement index of each skill of the person corresponding to the personnel file to be classified, the word vector corresponding to each word segmentation, the electronic information data and the skills required for the current task.

[0153] In this step, the skill mastery level of different skills of each personnel is evaluated in combination with the electronic information data of the personnel files to be classified.

[0154] In one embodiment, Figure 3As shown, according to the task requirement index of each skill of the personnel corresponding to the personnel files to be classified, the word vector corresponding to each word segmentation, the electronic information data and the skills required for the current task, the skill mastery level of each skill required for the current task of the personnel corresponding to each personnel file to be classified can be obtained by performing the following steps S1031-S1035 on the electronic information data of each personnel file to be classified and each skill required for the current task:

[0155] S1031. Calculate the cosine similarity between each text segmentation and the current skill keyword in the personal skill module in the current electronic information data based on the word vector corresponding to each text segmentation and the current electronic information data.

[0156] The current electronic information data is the electronic information data of the personnel file to be classified, the text segmentation represents the segmentation in the current electronic information data; the current skill keyword is the keyword corresponding to the current skill required for the current task.

[0157] For the needs of the current task, the task demand index of different skills can reflect the complexity of the skill in the process of completing the task. In order to classify the personnel corresponding to the personnel files to be classified according to the classification needs of the current task, it is first necessary to evaluate the various skill levels of the personnel corresponding to different personnel files.

[0158] The establishment of a personal profile can usually be divided into different modules such as basic personal information, personal skills, and work experience. For the skills required for the current task, the keywords of different skills are converted into word vector form through the same word2vec method as in the preprocessing process of step S101.

[0159] It is worth noting that the work experience module can also be replaced by a project experience module, or other modules. The present disclosure does not limit the specific name of the module, as long as the module can reflect the personnel's work experience or project manager.

[0160] The cosine similarity between each segmentation in the personal skill module of the personnel file corresponding to the current personnel to be classified and the keyword of the current skill is calculated, and the segmentation with the largest cosine similarity to the keyword of the current skill is called the target segmentation.

[0161] S1032. Obtain the mastery position coefficient of the current skill of the personnel corresponding to the current personnel file to be classified according to the personal skill module and the target segmentation in the current electronic information data; the target segmentation is the text segmentation corresponding to the maximum cosine similarity.

[0162] Select the current skill keyword with the largest cosine similarity The segmentation of is taken as the target segmentation, and then the position of the target segmentation in the segment of the personal skill module of the personnel file to be classified and the position of the segment in the personal skill module are calculated. When compiling the personal skill module in the file, the relatively basic skills that the personnel file is relatively familiar with are usually written in the front position to highlight the individual's advantages in previous work or study. Therefore, the position of the selected target segmentation can be used as one of the parameters for evaluating the mastery level of the current skills of the personnel file.

[0163] In one embodiment, Figure 4 As shown, according to the personal skill module and target segmentation in the current electronic information data, the mastery position coefficient of the current skill of the person corresponding to the current personnel file to be classified is obtained, including the following sub-steps S10321-S10323:

[0164] S10321. Obtain a first number of all text participles before a target participle in a target paragraph and a second number of all text participles included in the target paragraph; the target paragraph is the paragraph in the personal skill module where the target participle is located.

[0165] S10322. Obtain the number of previous paragraphs before the target paragraph in the personal skill module, and the total number of paragraphs in the personal skill module.

[0166] S10323. Obtain the mastery position coefficient of the current skill of the person corresponding to the personnel file to be classified according to the first number, the second number, the previous number of paragraphs and the total number of paragraphs.

[0167] The mastery position coefficient here represents the subjective mastery level of the skill.

[0168] Calculate the relative position of the selected target word x in the personal skill module of the personnel file to be classified, and the relative position of the paragraph to which the target word x belongs in the personal skill module. In the form of, under the dimension of the target segmentation, c represents the first number of all text segmentations before the target segmentation, d represents the second number of all text segmentations in the paragraph where the target segmentation is located, under the dimension of the paragraph, where c represents the number of previous paragraphs before the paragraph to which the target segmentation x belongs, and d represents the number of all paragraphs in the personal skill module. After obtaining the relative positions under the two dimensions, the two relative positions are multiplied to obtain P, and p represents the mastery position coefficient of the current skill. According to the above analysis, the smaller the p value, the better the mastery of the current skills by the personnel corresponding to the personnel profile to be classified.

[0169] S1033. According to the work experience module and target segmentation in the current electronic information data, the skill mastery degree and skill necessity degree of the current skill in each work experience in the work experience module are obtained.

[0170] However, the current skill mastery position coefficient can only indicate the individual relative situation of the skills mastered by the personnel corresponding to the personnel file to be classified. It fails to analyze the personnel's skill mastery level from a more objective standpoint. It is also necessary to conduct a comprehensive evaluation based on the actual work experience content corresponding to the target participle in the file.

[0171] When a personal profile describes work experiences of different complexity, the length of the text usually has certain differences: for a person's past work experience with high complexity, a relatively long length is usually used to describe it, and when the paragraph word segmentation vector difference is large, the amount of information used to describe the current experience is relatively large; in addition, work experience with high complexity usually requires the integration of multiple technologies, and the types of skill word segmentation involved are usually relatively more. In one embodiment, the above-mentioned work experience module and target word segmentation in the current electronic information data are used to obtain the skill mastery and skill necessity of the current skill in each work experience of the person corresponding to the current personnel profile to be classified, such as Figure 5 As shown, the following sub-steps S10331-S10337 are included:

[0172] S10331, obtaining a third number of all text segmentations included in each paragraph in the personal skill module;

[0173] S10332, calculating the information entropy between the word vectors of each word in each paragraph;

[0174] S10333. Obtain the degree of skill mastery reflected in each work experience based on the third quantity and information entropy.

[0175] Get the third number of text segmentations included in each paragraph in the personal skill module, and then calculate the information entropy of the word vector of each segmentation in each paragraph. According to the above analysis, by multiplying the third number of text segmentations by the information entropy, the degree of skill mastery reflected in each work experience can be obtained, and H is used as one of the parameters of the degree of skill mastery reflected in the current work experience.

[0176] In addition, in the process of evaluating different work experiences, it is also necessary to analyze the importance of various skills in different work experiences. The amount of description of skills can usually reflect the necessity of the current skills in this work experience: in a text sentence containing the current skill, when the number of target segmented words in the text sentence is more, it means that the description of the current skill in the current work experience is richer, and the necessity of the current skill in this work experience is usually relatively high; at the same time, for a text sentence containing the current skill, if the description of other skills is less, it means that the degree to which each text segmentation of the sentence describes the current skill is higher, and the importance of the current skill in the current work experience is higher.

[0177] S10334. Obtain the target sentence where the target word segment is located in each target work experience paragraph.

[0178] S10335. Obtain a fourth number of text segmentations contained in each target sentence.

[0179] S10336. Obtain the fifth number of text segmentations corresponding to other skill keywords contained in each target sentence.

[0180] S10337. Obtain the necessary degree of the current skill in each work experience based on the fourth quantity and the fifth quantity.

[0181] Calculate the fourth number g of all text segmentations contained in the sentence position of the target segmentation x in a certain work experience in the work experience module of the personnel file corresponding to the current personnel to be classified, and count the fifth number h of text segmentations corresponding to other skill keywords contained in the sentence.

[0182] According to the above analysis, calculate , where c represents the adjustment coefficient. In this example, c=1. Indicates the degree of necessity of the current skill corresponding to the target participle x in a certain work experience. According to the above analysis, the larger the value, the more necessary the current skill is in the current work experience.

[0183] S1034. According to the skill mastery level and skill necessity level, obtain the objective skill mastery level of the current skill of the personnel corresponding to the personnel file to be classified.

[0184] After obtaining the skill mastery level of the current skill corresponding to the target participle in the work experience and the skill necessity level in different work experiences, the objective skill mastery level K represented by the current skill corresponding to the target participle is evaluated. The specific calculation process can be expressed by the following formula:

[0185]

[0186] In the above formula, K represents the objective mastery level of the skill represented by the current skill corresponding to the target segmentation; m represents the number of work experiences involving a certain skill in the personnel file corresponding to the current personnel to be classified. represents the degree of necessity of the i-th skill in work experience, It represents the skill mastery level of the i-th skill reflected in the work experience, through It reflects the necessary degree of current skills in this work experience, and takes the weight as the weight to perform weighted summation of the skill mastery levels reflected in different work experiences, and uses K to represent the objective mastery level of current skills.

[0187] S1035. Obtain the skill mastery level of the current skill of the personnel corresponding to the personnel file to be classified according to the mastery position coefficient and the objective skill mastery level.

[0188] After obtaining the subjective skill mastery level and the objective skill mastery level of the current skill, combine them to calculate the skill mastery level of the current skill of the personnel corresponding to the personnel file to be classified , the specific calculation process can be expressed by the following formula:

[0189] ;

[0190] In the formula, Indicates the skill mastery level of the personnel corresponding to the personnel file to be classified; Indicates the maximum cosine similarity corresponding to the current skill, represents the objective mastery level of the skill represented by the target participle, Indicates the mastery position coefficient of the current skill.

[0191] According to the above analysis, by combining the subjective skill mastery level and the relatively objective skill mastery level parameters and adjusting the logical relationship, the cosine similarity between the skill keywords and the actual skill segmentation words is used to analyze the skill mastery level of the personnel corresponding to the current profile, reducing the impact of text differences on the analysis process. Indicates the skill mastery level of the current skill.

[0192] In one embodiment, it is also necessary to evaluate the learning ability of the personnel corresponding to the personnel files to be classified for various skills required for the current task. According to the task requirement index of each skill of the personnel corresponding to the personnel files to be classified, the word vector corresponding to each word segmentation, the electronic information data and the skills required for the current task, the learning ability of each personnel corresponding to the personnel files to be classified under the condition of the current task is obtained. The following steps S1036-S10311 can be performed on the electronic information data of each personnel file to be classified:

[0193] S1036. Obtain age data in the current electronic information data.

[0194] S1037. Obtain the sixth number of text segmentations included in the education experience part of the current electronic information data.

[0195] S1038. Obtain a first learning ability based on the age data and the sixth quantity.

[0196] When age data appears in the basic personal information module in the current electronic information data, the first learning ability can represent the learning ability embodied by the basic personal information.

[0197] In actual tasks, participants often need to constantly learn new knowledge to meet the needs of task completion. Therefore, in addition to the degree of mastery of different skills, the learning ability of participants is also one of the important indicators of file classification.

[0198] Taking the example that age data appears in the basic personal information module in the current electronic information data, and the educational experience of the participants also appears in the basic personal information module, the learning ability usually needs to be evaluated in combination with the age data in the personnel file, because for relatively young participants, their learning ability is usually stronger than that of older participants; at the same time, the educational experience of the participants in the basic personal information module is also one of the factors to determine their learning ability. The richer the educational experience, the richer the learning experience, and their learning ability is usually relatively strong; in the present disclosure, the age data in the basic personal information module in the electronic information data of the personnel file to be classified is obtained and used as the denominator, and the sixth number of text segmentations included in the education experience part is collected and used as the numerator to obtain a score Q, which represents the learning ability reflected by the basic personal information in the electronic information data of the personnel file to be classified.

[0199] S1039. Obtain the working hours for each working experience.

[0200] S10310. Calculate the learning ability reflected in each work experience based on the working hours of each work experience, the number of all skills included in each work experience, the necessary degree of all skills, and the objective mastery level of all skills.

[0201] In addition, in addition to basic personal information, learning ability can also be evaluated in combination with past work experience: for work experience that includes weak skill mastery, the more necessary the skill is in the project, the shorter the completion time is, which means that the person corresponding to the current personnel file to be classified needs to continue learning in the process of completing the project in order to complete the task efficiently.

[0202] Calculate the working time t of each work experience in the electronic information data of the personnel files to be classified, and calculate the learning ability reflected in each work experience based on the necessary degree of skills involved in the current work experience and the personnel's mastery of skills , where w represents the number of skills included in the current work experience, Indicates the working hours of the current job experience, represents the degree of skill necessity of the jth skill in work experience, represents the objective mastery level of the jth skill. Based on the above analysis, by combining the length of work experience and skill learning information, Indicates the learning ability reflected in current work experience.

[0203] S10311. Obtain the learning ability reflected by the current electronic information data based on the first learning ability and the learning abilities reflected by each work experience.

[0204] Then, after combining the learning ability reflected by the work experience with the first learning ability, U is used to represent the learning ability reflected by the current electronic information data, where The larger the value, the stronger the learning ability reflected by the archival information.

[0205] However, for different tasks, evaluating learning ability only through current electronic information data is likely to ignore the impact of different learning directions in the task, so further analysis is needed in combination with the actual needs of the task.

[0206] If the overall skill mastering structure is similar to the skill requirements of the current task, then it means that the skill structure mastered by the work and learning status of the personnel corresponding to the current personnel profile is relatively matched with the current task. For the personnel corresponding to the current personnel profile, their learning direction is relatively close to the skills required for the current task. When facing the skill learning requirements in the task, their learning speed is usually relatively fast, so the following steps S10312-S10315 need to be performed:

[0207] S10312. Sort the task requirement index corresponding to each skill according to a preset order to obtain a first sequence.

[0208] S10313. Sort the skill mastery levels of various skills in the current electronic information data according to a preset order to obtain a second sequence.

[0209] For example, the skills required for the current task are sorted into the first sequence according to the alphabetical order of the corresponding task requirement index, and then the corresponding skill mastery levels in the current electronic information data are sorted into the second sequence according to the same order.

[0210] S10314. Obtain the correlation coefficient between the first sequence and the second sequence.

[0211] S10315. Obtain the learning ability of the personnel corresponding to the personnel files to be classified under the current task conditions based on the learning ability and correlation coefficient reflected in the current electronic information data.

[0212] The correlation coefficient of the two sequences can be calculated, and the sigmoid function can be used to set the value range between [0,1], represented by r. According to the above analysis, the larger the value, the higher the degree of match between the skill mastery structure of the personnel corresponding to the personnel profile to be classified and the current task, and the speed of skill learning is usually relatively fast.

[0213] For the learning ability reflected in the archive information, the skill mastery structure of the personnel corresponding to the current personnel archive to be classified is adjusted to match the current task. It represents the learning ability of archivists under the current task conditions. The specific calculation process can be expressed by the following formula:

[0214] ;

[0215] In the formula, represents the learning ability of archivists under the current task conditions, r represents the correlation coefficient, Indicates the learning ability reflected by current electronic information data.

[0216] S104. Obtain the clustering distance of the electronic information data of every two personnel files to be classified according to the skills, skill mastering level and learning ability required for the current task.

[0217] The present disclosure also optimizes the personnel file data clustering distance measurement method, specifically, Figure 6 As shown, the following steps are performed on the electronic information data of every two personnel files to be classified, including the following steps S1041-S1043:

[0218] S1041. Obtain the same skills of the personnel corresponding to the two personnel files to be classified;

[0219] S1042. From the same skills of the personnel corresponding to the two personnel files to be classified, obtain a target skill included in the skills required for the current task;

[0220] S1043. Obtain the clustering distance of the electronic information data of the two personnel files to be classified according to the target skills, the skill mastery levels of the personnel corresponding to the two personnel files to be classified for each target skill, and the learning abilities of the personnel corresponding to the two personnel files to be classified.

[0221] For each personnel profile, the different skill mastering levels and learning abilities corresponding to the current task are combined into multidimensional data, and the data clustering distance D of the personnel profile to be classified is calculated. The specific calculation process can be expressed by the following formula:

[0222] ;

[0223] In the above formula, represents the data clustering distance between personnel files a to be classified and personnel files b to be classified, n represents the number of skills required for the current task, represents the task demand index for the i-th skill, represents the square of the difference between the skill mastery level of personnel profile a for the i-th skill and the skill mastery level of personnel profile b for the i-th skill, It represents the square of the difference between the learning ability corresponding to personnel profile a and the learning ability corresponding to personnel profile b.

[0224] S105, clustering the electronic information data of each personnel file to be classified according to the clustering distance.

[0225] The electronic information data of each personnel file to be classified is clustered. After obtaining the distance measurement method of the file data, the electronic information data of different personnel files can be clustered by the K-means clustering algorithm, in which the clustering parameter k is obtained by the elbow method. The clustering process combines different task requirements to improve the classification accuracy under the task requirements.

[0226] S106: Save the clustering results.

[0227] After clustering the electronic information data of personnel files by the above method, the ability of personnel in the same cluster to solve the current task is similar. Dispatching tasks based on the clustering results can reduce the cost of human resource decision-making. The files are classified according to the clustering results of the data operation module, and the classification results are saved in the memory.

[0228] This application evaluates the importance of the current task to different skill requirements by training a classification network; analyzes the skill mastery level of different personnel in the electronic information data of each personnel file to be classified, and analyzes the learning ability of the personnel corresponding to each personnel file to be classified in combination with the importance of the current task to each skill requirement, and further optimizes the clustering distance measurement method of the electronic information data of each personnel file to be classified under the current task requirements; finally, clustering is performed on the electronic information data of each personnel file to be classified. Compared with the existing methods, the clustering distance measurement process takes into account the skills required for the current task, the skill mastery level of each person corresponding to each personnel file to be classified for each skill required for the current task, and the learning ability of each person corresponding to each personnel file to be classified under the conditions of the current task. Since the clustering distance evaluation method between the electronic information data of each personnel file to be classified is optimized, the clustering result is more in line with the needs of the current task, thereby reducing the cost of human resource decision-making and improving planning efficiency.

[0229] Based on the same inventive concept, the embodiment of the present application also provides an electronic information data classification system for archive management for implementing the electronic information data classification method for archive management involved above. The implementation scheme for solving the problem provided by the system is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more embodiments of the electronic information data classification system for archive management provided below can refer to the limitations of the electronic information data classification method for archive management above, and will not be repeated here.

[0230] Figure 7 A structural block diagram of an electronic information data classification system method for archive management provided by an embodiment of the present invention, such as Figure 7 As shown, the system comprises:

[0231] The first acquisition module 11 is used to acquire the electronic information data of each personnel file to be classified, and pre-process the electronic information data to obtain the word vector corresponding to each word segment in the electronic information data;

[0232] The second acquisition module 12 is used to input the information of the current task and the electronic information data of the personnel file to be classified into the trained classification network, and output the task demand index of each skill of the personnel corresponding to the personnel file to be classified under the condition of the current task;

[0233] The third acquisition module 13 is used to obtain the skill mastery level of each person corresponding to the personnel file to be classified for each skill required for the current task and the learning ability of each person corresponding to the personnel file to be classified under the condition of the current task according to the task requirement index of each skill of the person corresponding to the personnel file to be classified, the word vector corresponding to each word segmentation, the electronic information data and the skills required for the current task;

[0234] A fourth acquisition module 14, for acquiring the clustering distance of the electronic information data of every two personnel files to be classified according to the skills required for the current task, the skill mastering level and the learning ability;

[0235] A clustering processing module 15, used for clustering the electronic information data of each of the personnel files to be classified according to the clustering distance;

[0236] The saving module 16 is used to save the clustering result.

[0237] In one embodiment, in preprocessing the electronic information data to obtain word vectors corresponding to each word segment in the electronic information data, the first acquisition module is specifically used to:

[0238] Performing word segmentation processing on the text in the electronic information data to obtain word segments corresponding to the text;

[0239] Removing stop words from each of the participles;

[0240] Performing Chinese word form restoration processing on each of the word segments after removing the stop words;

[0241] Each of the word segments after Chinese lemma restoration is converted into a word vector.

[0242] In one embodiment, before inputting the information of the current task and the electronic information data of the personnel file to be classified into the trained neural network, the second acquisition module is further used to:

[0243] Acquire a sample set, the sample set including: information of the current task, sample skill keywords of the field to which the preset personnel files to be classified correspond to the personnel files, and preset task requirement indexes of the marked sample skill keywords;

[0244] Dividing the sample set into a training set and a validation set;

[0245] Inputting the training set into the original classification network, iteratively training the original classification network using a gradient descent method, and updating the parameters in the original classification network using a preset loss function and the verification set;

[0246] When the preset loss function converges, the iterative training of the original classification network is stopped to obtain the trained classification network.

[0247] In one embodiment, in the aspect of acquiring the skill mastery level of each skill required for the current task of each person corresponding to the personnel file to be classified according to the task requirement index of each skill of the person corresponding to the personnel file to be classified, the word vector corresponding to each word segmentation, the electronic information data and the skills required for the current task, the third acquisition module is specifically used to:

[0248] The following steps are performed for each electronic information data of the personnel file to be classified and each skill required for the current task:

[0249] According to the word vector corresponding to each text segmentation and the current electronic information data, the cosine similarity between each text segmentation and the current skill keyword in the personal skill module in the current electronic information data is calculated; the current electronic information data is the electronic information data of the personnel file to be classified, and the text segmentation represents the segmentation in the current electronic information data; the current skill keyword is the keyword corresponding to the current skill required for the current task;

[0250] According to the personal skill module and the target segmentation in the current electronic information data, the mastery position coefficient of the current skill of the person corresponding to the current personnel file to be classified is obtained; the target segmentation is the text segmentation corresponding to the largest cosine similarity;

[0251] According to the work experience module in the current electronic information data and the target segmentation, obtaining the skill mastery degree and skill necessity degree of the current skill in each work experience in the work experience module;

[0252] According to the skill mastery degree and the skill necessity degree, obtaining the objective skill mastery level of the current skill of the person corresponding to the personnel file to be classified;

[0253] According to the mastering position coefficient and the objective mastering level of the skill, the skill mastering level of the current skill of the person corresponding to the personnel file to be classified is obtained.

[0254] In one embodiment, in the aspect of obtaining the mastery position coefficient of the current skill of the person corresponding to the current personnel file to be classified according to the personal skill module in the current electronic information data and the target segmentation, the third acquisition module is specifically used to:

[0255] Obtaining a first number of all the text segmentations before the target segmentation in a target paragraph and a second number of all the segmentations included in the target paragraph; the target paragraph is the paragraph where the target segmentation in the personal skill module is located;

[0256] Obtaining the number of previous paragraphs before the target paragraph in the personal skill module and the total number of paragraphs in the personal skill module;

[0257] The mastery position coefficient of the current skill of the person corresponding to the personnel file to be classified is obtained according to the first number, the second number, the previous number of paragraphs and the total number of paragraphs.

[0258] In one embodiment, in terms of obtaining the skill mastery level and skill necessity level of the current skill in each work experience of the work experience module and the target segmentation in the current electronic information data, the third acquisition module is specifically used to:

[0259] Obtaining a third number of all the text segmentations included in each paragraph in the work experience module;

[0260] Calculate the information entropy between the word vectors of each of the text segmentations in each paragraph in the work experience module;

[0261] According to the third quantity and the information entropy, obtaining the skill mastery level reflected by each of the work experiences;

[0262] Obtain the target paragraph in which the target word appears in the work experience module;

[0263] Obtaining the target sentence where the target word is located in each of the target paragraphs;

[0264] Obtaining a fourth number of the text segmentations contained in each of the target sentences;

[0265] Obtaining a fifth number of the text segmentations corresponding to the other skill keywords contained in each of the target sentences;

[0266] The skill necessity degree of the current skill in each of the work experiences is obtained based on the fourth quantity and the fifth quantity.

[0267] In one embodiment, in terms of acquiring the learning ability of each person corresponding to the personnel files to be classified under the condition of the current task according to the task requirement index of each skill of the person corresponding to the personnel files to be classified, the word vector corresponding to each word segmentation, the electronic information data and the skills required for the current task, the third acquisition module is specifically used to:

[0268] The following steps are performed for each electronic information data of the personnel file to be classified:

[0269] Acquire age data in the current electronic information data;

[0270] Acquire a sixth number of the text segmentations included in the education experience part of the current electronic information data;

[0271] acquiring a first learning ability according to the age data and the sixth quantity;

[0272] Obtain the working hours of each of the working experiences;

[0273] Calculate the learning ability reflected in each of the work experiences based on the working hours of each of the work experiences, the number of all the skills included in each of the work experiences, the necessary degree of all the skills, and the objective mastery level of all the skills;

[0274] Acquire the learning ability reflected by the current electronic information data according to the first learning ability and the learning abilities reflected by each of the work experiences;

[0275] Sorting the task requirement indexes corresponding to the skills in a preset order to obtain a first sequence;

[0276] Sorting the skill mastering level of each skill in the current electronic information data according to the preset order to obtain a second sequence;

[0277] Obtaining a correlation coefficient between the first sequence and the second sequence;

[0278] According to the learning ability reflected by the current electronic information data and the correlation coefficient, the learning ability of the personnel corresponding to the current personnel file to be classified under the condition of the current task is obtained.

[0279] In one embodiment, in the aspect of obtaining the clustering distance of the electronic information data of every two personnel files to be classified according to the skills, the skill mastering level and the learning ability, the fourth obtaining module is specifically used for:

[0280] The following steps are performed for each electronic information data of two personnel files to be classified:

[0281] Obtain the same skills of the personnel corresponding to the two personnel files to be classified;

[0282] From the same skills of the personnel corresponding to the two personnel files to be classified, obtaining a target skill included in the skills required for the current task;

[0283] The clustering distance of the electronic information data of the two personnel files to be classified is obtained according to the target skills, the skill mastery levels of the personnel corresponding to the two personnel files to be classified for each target skill and the learning abilities of the personnel corresponding to the two personnel files to be classified.

[0284] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0285] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0286] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for classifying electronic information data for archive management, characterized in that: The method comprises: Obtaining electronic information data of each personnel file to be classified, and preprocessing the electronic information data to obtain word vectors corresponding to each word segment in the electronic information data; Inputting the information of the current task and the electronic information data of the personnel file to be classified into the trained classification network, and outputting the task demand index of each skill of the personnel corresponding to the personnel file to be classified under the condition of the current task; According to the task requirement index of each skill of the personnel corresponding to the personnel files to be classified, the word vector corresponding to each word segmentation, the electronic information data and the skills required for the current task, the skill mastery level of each person corresponding to the personnel files to be classified for each skill required for the current task and the learning ability of each person corresponding to the personnel files to be classified under the condition of the current task are obtained; The method for obtaining the skill mastery level of each person corresponding to each of the personnel files to be classified for each skill required for the current task is as follows: The following steps are performed for each electronic information data of the personnel file to be classified and each skill required for the current task: According to the word vector corresponding to each text segmentation and the current electronic information data, the cosine similarity between each text segmentation and the current skill keyword in the personal skill module in the current electronic information data is calculated; the current electronic information data is the electronic information data of the personnel file to be classified, and the text segmentation represents the segmentation in the current electronic information data; the current skill keyword is the keyword corresponding to the current skill required for the current task; According to the personal skill module and the target segmentation in the current electronic information data, the mastery position coefficient of the current skill of the person corresponding to the current personnel file to be classified is obtained; the target segmentation is the text segmentation corresponding to the largest cosine similarity; According to the work experience module in the current electronic information data and the target segmentation, obtaining the skill mastery degree and skill necessity degree of the current skill in each work experience in the work experience module; According to the skill mastery degree and the skill necessity degree, obtaining the objective skill mastery level of the current skill of the person corresponding to the personnel file to be classified; According to the mastering position coefficient and the objective mastering level of the skill, obtaining the skill mastering level of the current skill of the person corresponding to the personnel file to be classified; The method for obtaining the learning ability of each person corresponding to each of the personnel files to be classified under the condition of the current task is: The following steps are performed for each electronic information data of the personnel file to be classified: Acquire age data in the current electronic information data; Acquire a sixth number of the text segmentations included in the education experience part of the current electronic information data; acquiring a first learning ability according to the age data and the sixth quantity; Obtain the working hours of each of the working experiences; Calculate the learning ability reflected in each of the work experiences based on the working hours of each of the work experiences, the number of all the skills included in each of the work experiences, the necessary degree of all the skills, and the objective mastery level of all the skills; Acquire the learning ability reflected by the current electronic information data according to the first learning ability and the learning abilities reflected by each of the work experiences; Sorting the task requirement indexes corresponding to the skills in a preset order to obtain a first sequence; Sorting the skill mastering level of each skill in the current electronic information data according to the preset order to obtain a second sequence; Obtaining a correlation coefficient between the first sequence and the second sequence; According to the learning ability reflected by the current electronic information data and the correlation coefficient, the learning ability of the personnel corresponding to the personnel file to be classified under the condition of the current task is obtained; According to the skills required for the current task, the skill mastering level and the learning ability, obtaining the clustering distance of the electronic information data of every two personnel files to be classified; Clustering the electronic information data of each of the personnel files to be classified according to the clustering distance; The clustering result is saved.

2. The electronic information data classification method for archive management according to claim 1, characterized in that: The preprocessing of the electronic information data to obtain a word vector corresponding to each word segment in the electronic information data includes: Performing word segmentation processing on the text in the electronic information data to obtain word segments corresponding to the text; Removing stop words from each of the participles; Performing Chinese word form restoration processing on each of the word segments after removing the stop words; Each of the word segments after Chinese lemma restoration is converted into a word vector.

3. The electronic information data classification method for archive management according to claim 1, characterized in that: Before inputting the information of the current task and the electronic information data of the personnel files to be classified into the trained neural network, the method further includes: Acquire a sample set, the sample set including: information of the current task, sample skill keywords of the field to which the preset personnel files to be classified correspond to the personnel files, and preset task requirement indexes of the marked sample skill keywords; Dividing the sample set into a training set and a validation set; Inputting the training set into the original classification network, iteratively training the original classification network using a gradient descent method, and updating the parameters in the original classification network using a preset loss function and the verification set; When the preset loss function converges, the iterative training of the original classification network is stopped to obtain the trained classification network.

4. The electronic information data classification method for archive management according to claim 1, characterized in that: The step of obtaining the mastering position coefficient of the current skill of the person corresponding to the current personnel file to be classified according to the personal skill module in the current electronic information data and the target segmentation includes: Obtaining a first number of all the text segmentations before the target segmentation in a target paragraph and a second number of all the segmentations included in the target paragraph; the target paragraph is the paragraph where the target segmentation in the personal skill module is located; Obtaining the number of previous paragraphs before the target paragraph in the personal skill module and the total number of paragraphs in the personal skill module; The mastery position coefficient of the current skill of the person corresponding to the personnel file to be classified is obtained according to the first number, the second number, the previous number of paragraphs and the total number of paragraphs.

5. The electronic information data classification method for archive management according to claim 4 is characterized in that: According to the work experience module in the current electronic information data and the target segmentation, the skill mastery degree and skill necessity degree of the current skill in each work experience in the work experience module are obtained, including: Obtaining a third number of all the text segmentations included in each paragraph in the work experience module; Calculate the information entropy between the word vectors of each of the text segmentations in each paragraph in the work experience module; According to the third quantity and the information entropy, obtaining the skill mastery level reflected by each of the work experiences; Obtain the target paragraph in which the target word appears in the work experience module; Obtaining the target sentence where the target word is located in each of the target paragraphs; Obtaining a fourth number of the text segmentations contained in each of the target sentences; Obtaining a fifth number of the text segmentations corresponding to the other skill keywords contained in each of the target sentences; The skill necessity degree of the current skill in each of the work experiences is obtained based on the fourth quantity and the fifth quantity.

6. The electronic information data classification method for archive management according to claim 1, characterized in that: The step of obtaining the clustering distance of the electronic information data of every two personnel files to be classified according to the skills, the skill mastering level and the learning ability includes: The following steps are performed for each electronic information data of two personnel files to be classified: Obtain the same skills of the personnel corresponding to the two personnel files to be classified; From the same skills of the personnel corresponding to the two personnel files to be classified, obtaining a target skill included in the skills required for the current task; The clustering distance of the electronic information data of the two personnel files to be classified is obtained according to the target skills, the skill mastery levels of the personnel corresponding to the two personnel files to be classified for each target skill and the learning abilities of the personnel corresponding to the two personnel files to be classified.

7. An electronic information data classification system for archive management, characterized in that: The system comprises: A first acquisition module is used to acquire the electronic information data of each personnel file to be classified, and pre-process the electronic information data to obtain the word vector corresponding to each word segment in the electronic information data; The second acquisition module is used to input the information of the current task and the electronic information data of the personnel file to be classified into the trained classification network, and output the task demand index of each skill of the personnel corresponding to the personnel file to be classified under the condition of the current task; The third acquisition module is used to obtain the skill mastery level of each person corresponding to the personnel file to be classified for each skill required for the current task and the learning ability of each person corresponding to the personnel file to be classified under the condition of the current task according to the task requirement index of each skill of the person corresponding to the personnel file to be classified, the word vector corresponding to each word segmentation, the electronic information data and the skills required for the current task; The method for obtaining the skill mastery level of each person corresponding to each of the personnel files to be classified for each skill required for the current task is as follows: The following steps are performed for each electronic information data of the personnel file to be classified and each skill required for the current task: According to the word vector corresponding to each text segmentation and the current electronic information data, the cosine similarity between each text segmentation and the current skill keyword in the personal skill module in the current electronic information data is calculated; the current electronic information data is the electronic information data of the personnel file to be classified, and the text segmentation represents the segmentation in the current electronic information data; the current skill keyword is the keyword corresponding to the current skill required for the current task; According to the personal skill module and the target segmentation in the current electronic information data, the mastery position coefficient of the current skill of the person corresponding to the current personnel file to be classified is obtained; the target segmentation is the text segmentation corresponding to the largest cosine similarity; According to the work experience module in the current electronic information data and the target segmentation, obtaining the skill mastery degree and skill necessity degree of the current skill in each work experience in the work experience module; According to the skill mastery degree and the skill necessity degree, obtaining the objective skill mastery level of the current skill of the person corresponding to the personnel file to be classified; According to the mastering position coefficient and the objective mastering level of the skill, obtaining the skill mastering level of the current skill of the person corresponding to the personnel file to be classified; The method for obtaining the learning ability of each person corresponding to each of the personnel files to be classified under the condition of the current task is: The following steps are performed for each electronic information data of the personnel file to be classified: Acquire age data in the current electronic information data; Acquire a sixth number of the text segmentations included in the education experience part of the current electronic information data; acquiring a first learning ability according to the age data and the sixth quantity; Obtain the working hours of each of the working experiences; Calculate the learning ability reflected in each of the work experiences based on the working hours of each of the work experiences, the number of all the skills included in each of the work experiences, the necessary degree of all the skills, and the objective mastery level of all the skills; Acquire the learning ability reflected by the current electronic information data according to the first learning ability and the learning abilities reflected by each of the work experiences; Sorting the task requirement indexes corresponding to the skills in a preset order to obtain a first sequence; Sorting the skill mastering level of each skill in the current electronic information data according to the preset order to obtain a second sequence; Obtaining a correlation coefficient between the first sequence and the second sequence; According to the learning ability reflected by the current electronic information data and the correlation coefficient, the learning ability of the personnel corresponding to the current personnel files to be classified under the condition of the current task is obtained; the fourth acquisition module is used to obtain the clustering distance of the electronic information data of every two personnel files to be classified according to the skills required for the current task, the skill mastery level and the learning ability; A clustering processing module, used for clustering the electronic information data of each of the personnel files to be classified according to the clustering distance; A saving module is used to save the clustering result.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the electronic information data classification method for archive management described in any one of claims 1 to 6 are implemented.

Images