A position screening method and system based on big data

Abstract

The application provides a position screening method based on big data, comprising the following steps: obtaining weight coefficients and index values of each evaluation target of a preselected position; obtaining a first score of the preselected position according to the sum of the products of the weight coefficients and the index values of the evaluation targets; obtaining evaluation data of a company to which the preselected position belongs, and obtaining frequency weights and word base numbers of each word in the evaluation data; obtaining a second score value of the preselected position according to the sum of the products of the word base numbers and the frequency weights of each word; the sum of the first score value and the second score value is a final score of the preselected position, and the preselected position with the highest final score is screened out. The application aims to solve the problem that a job seeker is difficult to determine the optimal job opportunity from multiple acceptance letters.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field, and particularly relates to a position screening method and system based on big data. BACKGROUND

[0002] Compared with traditional offline recruitment, online recruitment has great advantages in convenience and information transparency. Therefore, in the job-hunting process, a job seeker will encounter multiple positions that meet the conditions. When the job seeker receives multiple acceptance letters, it is usually difficult to determine which job is most suitable for oneself, and indecision will lead to missing an excellent job.

[0003] The prior art discloses a post or resume screening method, which screens the positions that meet the conditions according to the job positions, salaries and commuting times of job seekers, or screens the resumes that meet the conditions according to the recruitment information of recruiters. The screening method has few limited conditions and all the conditions are hard conditions, which leads to an inflexible screening result and is only suitable for screening a part of recruitment information from huge recruitment data, and is not suitable for scoring and screening multiple acceptance letters. Therefore, there is an urgent need for a job scoring method to help job seekers determine the optimal job opportunity from multiple positions. SUMMARY

[0004] The present application aims to at least solve the technical problems existing in the prior art, and provides a position screening method and system based on big data to solve the problem that a job seeker cannot determine the optimal job opportunity from multiple acceptance letters.

[0005] In order to achieve the above-mentioned purpose of the present application, according to a first aspect of the present application, a position screening method based on big data is provided, comprising the following steps: obtaining the weight coefficient and index value of each evaluation target of a preselected position; obtaining a first score of the preselected position according to the sum of the products of the weight coefficients and index values of the evaluation targets; obtaining evaluation data of a company to which the preselected position belongs, and obtaining the frequency weight and word base of each word in the evaluation data; obtaining a second score value of the preselected position according to the sum of the products of the word base and frequency weight of each word; the sum of the first score value and the second score value is the final score of the preselected position, and the preselected position with the highest final score is screened out.

[0006] Further, the generation step of the weight parameter is as follows: the position is divided into different evaluation targets, the comparative score of each evaluation target is obtained according to the scale method, and the weight coefficient of the evaluation target is obtained through the weight formula and the comparative score.

[0007] Further, the weight coefficient comprises a primary weight coefficient and a combined weight coefficient; the generation of the primary weight coefficient is as follows: the position is decomposed into evaluation targets of different levels to form an evaluation target tree diagram; the contrast score of the evaluation target in the same level is obtained according to the scale method, and the initial weight coefficient of the evaluation target is obtained according to the contrast score in the same level and the weight formula; the generation of the combined weight coefficient is as follows: the initial weight parameter of the evaluation target is multiplied by the initial weight parameter of the evaluation target in the previous level to obtain the combined weight coefficient of the evaluation target.

[0008] Further, the weight formula is as follows: wherein m is the total number of evaluation targets in the same level, a nm is the contrast score of the nth evaluation target and the mth evaluation target, a n1 is the contrast score of the nth evaluation target and the first evaluation target, a n1 is the contrast score of the nth evaluation target and the second evaluation target, W n is the intermediate weight value of the nth evaluation target; wherein n and m are positive integers.

[0009] Further, the generation of the index value is as follows: the information of each evaluation target of the preselected position is collected, the collected information is preprocessed to obtain the index value of each evaluation target, and the corresponding index collection table is generated.

[0010] Further, the evaluation data is obtained as follows: the initial evaluation of the company of the preselected position is obtained, and the initial evaluation is preprocessed to obtain the evaluation data.

[0011] Further, the frequency weight is obtained as follows: the evaluation data is subjected to word segmentation processing to obtain a plurality of words, and the TF-IDF (term frequency-inverse document frequency) algorithm is used to calculate the frequency weight of the words in the evaluation data.

[0012] Further, the step of obtaining the second score value of the preselected position according to the sum of the product of the word base and the frequency weight of each word is as follows: N words are selected from high to low according to the frequency weight, the product of the frequency weight and the word base of the N words is added to obtain the second score value, wherein N is a positive integer.

[0013] Further, the words comprise positive words, negative words and neutral words; the word base of the positive words is set to be a positive number, the word base of the negative words is set to be a negative number, and the word base of the neutral words is set to be 0.

[0014] In order to achieve the above-mentioned purpose of the present application, according to a second aspect of the present application, a big data-based position screening system is provided, which uses any of the above-mentioned big data-based position screening methods during operation, and comprises an acquisition module, a storage module, a decomposition module, a calculation module and an output module; the acquisition module is used for acquiring a preselected position, an index value and evaluation data; the storage module is used for storing an index collection table, an evaluation target tree diagram, a weight parameter of a rating target and a word base; the decomposition module is used for decomposing the evaluation data into a plurality of words; the calculation module is used for calculating a frequency weight of the words, a first score value, a second score value and a final score of the preselected position; and the output module is used for outputting the final score of the preselected position.

[0015] The technical principle of the present application is as follows: the present application decomposes a position into a plurality of target evaluations, scores the preselected position according to the index value and the weight parameter of each target evaluation, and if the weight parameter of the target evaluation is high and the index value is high, a first score value is high; secondly, the evaluation data of the collecting company is acquired, the words in the evaluation data are extracted, the evaluation of the company is scored by judging the word base and the frequency weight of the words, and if all the words are positive words and the word proportion is high, it means that the company is highly praised, and the second score value is high. Finally, the final score is obtained by the first score value and the second score value, which helps the user to screen the best position.

[0016] The beneficial effects of the present application are as follows: the present application mainly scores the position from the position condition and the company evaluation. The job seeker can set the weight coefficient of each evaluation target according to his / her own preference and development, so that the first score value is consistent with the subjective evaluation of the user on the position condition. Compared with the prior art, the scoring conditions of the present application can be adjusted, and the flexibility of the screening result is improved. The judgment accuracy of the position condition can be improved by increasing the number of evaluation targets. At the same time, the present application acquires and analyzes the company evaluation data through big data, which helps the job seeker to screen the high-quality company. The present application comprehensively analyzes and evaluates the position, helps the job seeker to quickly screen the position with high satisfaction and belonging to a high-quality company from a plurality of acceptance letters, and avoids the job seeker to miss the excellent job. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 is a step schematic diagram of a big data-based position screening method of the present application;

[0018] Figure 2 is a structure schematic diagram of an evaluation target tree diagram in the present application;

[0019] Figure 3 is a structure schematic diagram of a big data-based position screening system of the present application. DETAILED DESCRIPTION

[0020] Embodiments of the present application are described below in detail with reference to the accompanying drawings, wherein the same or similar components or components having the same or similar functions are denoted by the same or similar reference numerals throughout. The embodiments described below by reference to the accompanying drawings are exemplary and are intended only for the purpose of explaining the present application, and should not be understood as limiting the present application.

[0021] In the description of the present application, it should be understood that the terms "longitudinal", "transverse", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship shown in the drawings, and are only for the purpose of facilitating the description of the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore should not be understood as limiting the present application.

[0022] In the description of the present application, unless otherwise specified and limited, it should be noted that the terms "mounting", "connection", "connection" should be understood broadly, for example, it can be a mechanical connection or an electrical connection, it can be a communication between two elements, it can be a direct connection, or an indirect connection through an intermediate medium, and the specific meaning of the above terms can be understood by a person skilled in the art according to the specific circumstances.

[0023] As shown in the accompanying drawings, Figure 1 The present application provides a position screening method based on big data, comprising the following steps:

[0024] The weight coefficient and the index value of each evaluation target of the preselected position are obtained, the first score of the preselected position is obtained according to the sum of the product of the weight coefficient and the index value of the evaluation target, the evaluation data of the company to which the preselected position belongs is obtained, the frequency weight and the word base of each word in the evaluation data are obtained, the second score value of the preselected position is obtained according to the sum of the product of the word base and the frequency weight of each word, the sum of the first score value and the second score value is the final score of the preselected position, and the preselected position with the highest final score is screened out.

[0025] Preferably, the weight parameter generation step is as follows: the position is decomposed into different evaluation targets, the comparative score of each evaluation target is obtained according to the scale method, and the weight coefficient of the evaluation target is obtained through the weight formula and the comparative score. Preferably, the weight coefficient includes a primary weight coefficient and a combined weight coefficient:

[0026] The generation step of the primary weight coefficient is as follows: the position is decomposed into evaluation targets of different levels to form an evaluation target tree diagram; the comparative score of the evaluation target at the same level is obtained according to the scale method, and the initial weight coefficient of the evaluation target is obtained according to the comparative score at the same level and the weight formula;

[0027] Specifically, to precisely define the importance and impact of each element in the pre-selected position, the position is broken down into first-level evaluation objectives: compensation and benefits, personal development, and work environment; compensation and benefits are further broken down into second-level evaluation objectives: annual salary, social insurance and housing fund contributions, and shareholder options; personal development is further broken down into third-level evaluation objectives: team size, training frequency, and annual promotion frequency; and the work environment is further broken down into fourth-level evaluation objectives: including commuting time, working hours, and overtime hours. An evaluation objective tree diagram is generated based on these objectives, as shown in the figure below. Figure 2 As shown;

[0028]

[0029] Obtained by scaling Figure 2 The comparative scores of compensation and benefits, personal development, and work environment are shown in Table 1 using the scaling method. Based on the comparative scores, a judgment and optimization matrix for the first-level evaluation objectives is established, as shown in Table 2.

[0030]

[0031]

[0032] The initial weight coefficients of the evaluation targets are obtained based on the weighting formula and the comparative scores; the weighting formula is as follows:

[0033]

[0034] Where m is the total number of evaluation targets at the same level, a nm Let a be the comparison score between the nth evaluation objective and the mth evaluation objective. n1 Let a be the comparison score between the nth evaluation objective and the 1st evaluation objective. n1 W is the comparison score between the nth evaluation objective and the second evaluation objective. n Let m be the intermediate weight value of the nth evaluation target; where n and m are both positive integers, and n ≤ m.

[0035] Depend on Figure 2 It is known that the total number of evaluation objectives in the first level is m1 = 3. Based on the weighting formula and comparative scores, the median weight value of compensation and benefits is obtained. Median weighting of the work environment Median weighting for personal development

[0036] According to the normalization formula The intermediate weight values ​​are normalized to obtain the initial weight parameters W′. n Specifically, the initial weighting parameters for compensation and benefits. Initial weight parameters for personal development Initial weight parameters of the working environment

[0037] The comparison scores of the second-level evaluation objectives are obtained using the scaling method. Based on these comparison scores, a judgment and optimization matrix for the second-level evaluation objectives is established.

[0038] As shown in Table 3:

[0039]

[0040] Depend on Figure 2 From Table 3, we know that the total number of rating targets in the second level is m2 = 3. Following the calculation process of the initial weight parameters described above, we obtain the initial weight parameter W′ for the annual salary. 11 =0.6863; Initial weighting parameter W′ for the five social insurances and one housing fund 12 = 0.1634, the initial weight parameter W′ of the shareholder option 13 =0.1503;

[0041] The comparison scores of the third-level evaluation objectives are obtained using the scaling method. Based on these comparison scores, a judgment and optimization matrix for the third-level evaluation objectives is established, as shown in Table 4.

[0042]

[0043] Depend on Figure 2 From Table 4, we know that the total number of evaluation targets in the third level is m3 = 3. Following the calculation process of the initial weight parameters described above, we obtain the initial weight parameter W′ for the number of team members. 21 =0.6863; Initial weight parameter W′ for training sessions 22 =0.1634, the initial weight parameter W′ for annual promotion frequency 23 =0.1503;

[0044] The comparison scores of the fourth-level evaluation objectives are obtained using the scaling method. Based on these comparison scores, a judgment and optimization matrix for the fourth-level evaluation objectives is established.

[0045] As shown in Table 5:

[0046]

[0047] Depend on Figure 2 From Table 5, we know that the total number of rating targets in the fourth level is m4 = 3. Following the calculation process of the initial weight parameters described above, we obtain the initial weight parameter W′ for commuting time. 31 =0.1338; Initial weight parameter W′ for working hours32 = 0.4586, initial weight parameter W' of overtime length 33 = 0.4076

[0048] The generation step of the combined weight coefficient is as follows: multiplying the initial weight parameter of the evaluation target by the initial weight parameter of the evaluation target of the upper level to obtain the combined weight coefficient of the evaluation target.

[0049] Specifically, the calculation process of the combined weight coefficient is as follows: from the above formula (1), the initial weight parameter of the evaluation target is multiplied by the initial weight parameter of the evaluation target of the upper level to obtain the combined weight coefficient of the evaluation target. Figure 2 As shown in the above formula (1), the upper level evaluation target of the salary benefit, personal development and working environment is the position, the initial weight coefficient of the position is 1, then the combined weight parameter W"1 of the salary benefit is W'1x1=0.6370; the combined weight parameter W"2 of the personal development is W'2x1=0.1047; the combined weight parameter W"3 of the working environment is W'3x1=0.2583; the combined weight parameter W"4 of the annual salary is W'4xW'1=0.4371731; the combined weight parameter W"5 of the five insurances and one fund is W'5xW'1=0.1040858; the combined weight parameter W"6 of the shareholder option is W'6xW'1=0.0957411; the combined weight parameter W"7 of the team number is W'7xW'2=0.17727129; the combined weight parameter W"8 of the training frequency is W'8xW'2=0.04220622; the combined weight parameter W"9 of the annual promotion frequency is W'9xW'2=0.03882249; the combined weight parameter W"10 of the commuting length is W'10xW'3=0.01400886; the combined weight parameter W"11 of the working length is W'11xW'3=0.04801542; the combined weight parameter W"12 of the overtime length is W'12xW'3=0.04267572. 11 11 12 12 13 13 21 21 22 22 23 23 31 31 32 32 33 33

[0050] The generation step of the index value is as follows:

[0051] ​​​​​​​​​​​​​​​​​​The information of each evaluation target of the pre-selected position is collected, the collected information is pre-processed, the index value of each evaluation target is obtained, and the corresponding index collection table is generated. Specifically, if the collected information of the annual salary is 100000, then 100000 is the index value of the annual salary of the pre-selected position; if the collected information of the team size is 200, then 200 is the index value of the team size of the pre-selected position; if the collected information of the commuting time is 0.8 hours, then the index value of the commuting time is -0.8. The first score value is obtained by adding the product of the combination weight parameter of each evaluation target and the index value.

[0052] The evaluation data is obtained as follows:

[0053] The initial evaluation of the pre-selected position of the company is obtained, and the initial evaluation is pre-processed to obtain the evaluation data. Preferably, the initial evaluation includes the text evaluation content of the job seeker on the company on the Internet platform.

[0054] Specifically, the initial evaluation of the pre-selected position of the company is obtained by searching on the Internet. Since the information or data obtained by searching on the Internet includes a lot of redundant and useless information and chaotic data, in order to improve the quality of the obtained information or data, the initial evaluation is pre-processed to obtain the evaluation data of the company. Specifically, the pre-processing steps include data cleaning, data integration, data transformation and data reduction, and also include necessary processing such as auditing, screening and sorting before classifying or grouping the obtained data or information. Since the data processing method is prior art, it will not be described here.

[0055] The frequency weight is obtained as follows:

[0056] The evaluation data is processed to obtain a plurality of words, and the TF-IDF (Term Frequency-Inverse Document Frequency) algorithm is used to calculate the frequency weight of the words in the evaluation data.

[0057] Specifically, the words are obtained by processing the evaluation data by using the Jieba segmentation method, which can accurately segment sentences and has fast segmentation speed. The frequency weight of the words in the evaluation data is calculated by using the TF-IDF (Term Frequency-Inverse Document Frequency) algorithm. The calculation process of the frequency weight is as follows:

[0058] The frequency TF of the words in the evaluation data is calculated according to the word frequency formula ij ; the word frequency formula is as follows:

[0059]

[0060] wherein n i,j is the number of times the word appears in the evaluation data, and k n k,j is the total number of times all words appear in the evaluation data.

[0061] The inverse document frequency IDF of a word is calculated according to the inverse document frequency formula i The inverse document frequency formula is as follows:

[0062]

[0063] wherein IDF i is the inverse document frequency, |D| is the set of evaluation data of all companies; |{j: t i ∈d j} | represents the total number of evaluation data containing the word t i , i.e. the number of evaluation data with n i,j ≠ 0.

[0064] The frequency weight = TF ij × IDF i . The higher the frequency of a word in a certain evaluation data and the lower the frequency of the word in the set of evaluation data, the higher the frequency weight obtained, so the stronger the word's ability to predict the theme, and the greater the weight, and vice versa.

[0065] Preferably, the words include positive words, negative words and neutral words; the word base of a positive word is set as a positive number, the word base of a negative word is set as a negative number, and the word base of a neutral word is set as 0.

[0066] In order to accurately describe the company evaluation data using the score, the words are divided into positive words, negative words or neutral words according to the emotional color; the word base of a positive word is set as a positive number, the word base of a negative word is set as a negative number, and the word base of a neutral word is set as 0, and different positive word bases or negative word bases can also be set according to the emotional color density of the positive words or negative words, for example, the word base of the word "quasi-time off work" is set as +1, the word base of the word "fast promotion" is set as +2, the word base of the word "serious overtime" is set as -2, and the word base of the word "delayed wages" is set as -3, etc.

[0067] In the above manner, when the frequency weight of a positive word in the evaluation data is high, the second score value is higher, and when the frequency weight of a negative word in the company evaluation data is high, the second score value is lower. The second score value can accurately reflect the public's evaluation of the company, so as to help job seekers select good companies for evaluation.

[0068] The second score value of the pre-selected position is obtained according to the sum of the product of the term base and the frequency weight of each term.

[0069] The N terms are selected according to the frequency weight from high to low, and the second score value is obtained by adding the product of the frequency weight and the term base of the N terms, wherein N is a positive integer. For example, N=3, that is, the three terms with the highest weight are selected, and the product of the term base and the frequency weight of the three terms is added to obtain the second score value.

[0070] The final score of the pre-selected position is obtained by adding the first score value and the second score value, and the final scores of the plurality of pre-selected positions are sorted, so that the optimal position can be obtained from the pre-selected positions.

[0071] As shown in the accompanying Figure 3 The present application also provides a position screening system based on big data, which uses the position screening method based on big data described above during operation. The system comprises an acquisition module, a storage module, a decomposition module, a calculation module and an output module. The acquisition module is used to acquire pre-selected positions, index values and evaluation data. The storage module is used to store index collection tables, evaluation target tree graphs, weight parameters of rating targets and term bases. The decomposition module is used to decompose the evaluation data into a plurality of terms. The calculation module is used to calculate the frequency weight of the terms, the first score value, the second score value and the final score of the pre-selected positions. The output module is used to output the final score of the pre-selected positions.

[0072] In the description of the present specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example" or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0073] Although the embodiments of the present application have been shown and described, those skilled in the art can understand that various changes, modifications, replacements and variations can be made to the embodiments without departing from the principles and purposes of the present application, and the scope of the present application is defined by the claims and their equivalents.

Claims

1. A big data based job screening method, characterized in that, The method comprises the following steps: obtaining the weight coefficient and the index value of each evaluation target of the preselected position; obtaining the first score of the preselected position according to the sum of the product of the weight coefficient and the index value of the evaluation target; obtaining the evaluation data of the company to which the preselected position belongs, and obtaining the frequency weight and the word base of each word in the evaluation data; obtaining the second score value of the preselected position according to the sum of the product of the word base and the frequency weight of each word; the sum of the first score value and the second score value is the final score of the preselected position, and the preselected position with the highest final score is screened out.

2. The big data based job screening method as claimed in claim 1, wherein, The generation steps of the weight parameter are as follows: The position is divided into different evaluation targets, the comparative score of each evaluation target is obtained according to the scaling method, and the weight coefficient of the evaluation target is obtained through the weight formula and the comparative score. 3.The big data-based job screening method of claim 1 or 2, wherein, The weight coefficient includes a primary weight coefficient and a combined weight coefficient. The generation steps of the primary weight coefficient are as follows: the position is divided into different levels of evaluation targets to form an evaluation target tree diagram; the comparative score of the evaluation target at the same level is obtained according to the scaling method, and the initial weight coefficient of the evaluation target is obtained according to the comparative score at the same level and the weight formula; The generation steps of the combined weight coefficient are as follows: the initial weight parameter of the evaluation target is multiplied by the initial weight parameter of the evaluation target at the previous level to obtain the combined weight coefficient of the evaluation target.

4. The big data based job screening method as claimed in claim 3, wherein, The weight formula is as follows: wherein m is the total number of evaluation targets in the same level, a nm is the comparison score of the nth evaluation target and the mth evaluation target, a n1 is the comparison score of the nth evaluation target and the 1st evaluation target, a n1 is the comparison score of the nth evaluation target and the 2nd evaluation target, W n is the intermediate weight value of the nth evaluation target; wherein n and m are both positive integers.

5. The big data based job screening method of claim 1, 2 or 4, wherein, The generation steps of the index value are as follows: The information of each evaluation target of the preselected position is collected, the collected information is preprocessed, the index value of each evaluation target is obtained, and the corresponding index collection table is generated.

6. The big data based job screening method of claim 1, 2 or 4, wherein, The evaluation data is obtained as follows: The initial evaluation of the company of the preselected position is obtained, and the initial evaluation is preprocessed to obtain the evaluation data.

7. The big data based job screening method of claim 1, 2 or 4, wherein, The frequency weight is obtained as follows: The evaluation data is segmented to obtain a plurality of words, and the frequency weight of the word in the evaluation data is calculated using the TF-IDF (Term Frequency-Inverse Document Frequency) algorithm.

8. The big data based job screening method as claimed in claim 7, wherein, The steps of obtaining the second score value of the preselected position according to the sum of the product of the word base and the frequency weight of each word are as follows: N words are selected from high to low according to the frequency weight, and the product of the frequency weight and the word base of the N words is added to obtain the second score value, wherein N is a positive integer.

9. The big data based job screening method of claim 1, 2, 4, or 8, wherein: The words include positive words, negative words and neutral words; the word base of the positive words is set to a positive number, the word base of the negative words is set to a negative number, and the word base of the neutral words is set to 0. 10.A big data based job screening system, characterized in that, In the running process, the position screening method based on big data in any one of claims 1-9 is used, and the system comprises an acquisition module, a storage module, a decomposition module, a calculation module and an output module; The acquisition module is used to acquire the preselected position, the index value and the evaluation data; The storage module is used to store the index collection table, the evaluation target tree diagram, the weight parameter of the evaluation target and the word base; The decomposition module is used to decompose the evaluation data into a plurality of words; The calculation module is used to calculate the frequency weight of the word, the first score value, the second score value and the final score of the preselected position; The output module is used to output the final score of the preselected position.

Images