Search

CN-122022325-A - Personnel allocation scheme generation method and system based on multi-person decision tracing and conflict resolution

CN122022325ACN 122022325 ACN122022325 ACN 122022325ACN-122022325-A

Abstract

The invention relates to the technical field of intelligent allocation of enterprise personnel, and particularly discloses a personnel allocation scheme generation method and system based on multi-person decision tracing and conflict resolution. The method comprises the core steps of firstly constructing a personnel allocation decision tree containing time sequence characteristics, wherein each decision node records complete allocation instructions, participation decision personnel information and decision time stamps, secondly calculating comprehensive decision weight coefficients according to multi-dimensional factors such as the job level weight, emotion state analysis, historical decision effectiveness and association degree with candidate personnel of the decision personnel when multiple personnel are involved in decision, then integrating opinion of all parties through a conflict resolution model to generate an optimal personnel allocation scheme, and finally storing a complete decision process by utilizing a time sequence database to support decision tracing and scheme optimization. The invention provides a scientific and quantitative solution to the problem of opinion conflict in a multi-person decision scene, and the decision making efficiency and the scheme quality of enterprise personnel allocation are obviously improved.

Inventors

  • SUN ZHIXIN
  • Ran Xining
  • XU YUHUA

Assignees

  • 南京邮电大学

Dates

Publication Date
20260512
Application Date
20260129

Claims (10)

  1. 1. A personnel allocation scheme generation method based on multi-person decision tracing and conflict resolution is characterized by comprising the following steps: S1, receiving a personnel allocation request, wherein the personnel allocation request comprises target post information, the number of required personnel and capability requirements; S2, screening candidate persons meeting the conditions from a talent database to generate an initial candidate set; s3, constructing a multi-person decision interface for a plurality of decision makers to submit respective personnel allocation schemes; S4, aiming at the scheme submitted by each decision maker, calculating a decision maker weight coefficient based on the decision maker job level, the emotion state, the historical decision making effectiveness and the association degree of candidate personnel; s5, integrating the multiparty opinion through a conflict resolution model based on the decision maker weight coefficient to generate a comprehensive personnel allocation scheme; and S6, saving the decision process as a node into a decision tree structure, wherein the node comprises decision time, a participator, scheme details and a final determination scheme.
  2. 2. The method according to claim 1, wherein in step S3, the constructing a multi-person decision interface further comprises: s2.1, providing a visual interface to display the matching degree of the capability portraits and posts of the candidate personnel; S2.2, enabling each decision maker to independently select or sort candidate personnel; s2.3, recording the operation track and the decision time point of each decision maker.
  3. 3. The method according to claim 1, wherein in step S4, the decision maker weight coefficients are calculated using the following formula: Wherein, the Is the first The comprehensive weight system of the bit decision maker; A job level weight based on job level normalization; the value interval is 0, 1; the emotion state coefficient based on emotion analysis has higher positive emotion value; a risk assessment score for a decision maker's historical decisions; the association coefficient of the decision maker and the candidate person is obtained; , , , for the adjustable weight parameter of each dimension, satisfy 。
  4. 4. A method according to claim 3, wherein the emotional state coefficients The acquisition of (1) comprises: S4.1, analyzing text input emotion tendencies of a decision maker in a decision process through a natural language processing technology; s4.2, establishing a correlation model of the decision maker emotion state and decision quality according to the historical data; s4.3, calculating an emotion state coefficient based on the current decision text and the correlation model.
  5. 5. The method according to claim 1, wherein in step S5, the conflict resolution model employs a variable precision dominant coarse thresholding method, comprising: s5.1, constructing decision opinions of all parties into a decision table, wherein each row represents the evaluation of a decision maker on a candidate person; s5.2, defining conflict relations based on the advantage relations, and identifying contradictions among different decision opinions; s5.3, calculating a minimum conflict set, and determining a decision threshold; S5.4, generating an integrated optimal personnel allocation scheme based on the threshold value.
  6. 6. The method according to claim 1, wherein step S5 specifically comprises: Firstly, quantifying the dominant relationship of decision opinion and setting up a common Bit decision maker The weight vector is as follows Is directed to Position candidate Each decision maker For candidates The evaluation score of (2) is recorded as Candidate person Relative to Global dominance of (a) The definition is as follows: Wherein, the For the indication function, the condition in brackets is true 1, otherwise is 0; expressed in all decision maker weighted opinions, consider Is superior to Is a ratio of (3); Then, a conflict is identified and a conflict degree is calculated, and a conflict matrix is defined Its elements Representing candidates And The degree of conflict of the relative merits of (a): When (when) When the collision degree approaches 0.5, the average enemy of the number of people and the weight of the two supported parties is shown, the collision degree approaches to the maximum value 1, when When approaching 0 or 1, the conflict degree approaches to the minimum value 0, the system sets a conflict threshold value When (when) At the time, mark Is a conflict pair; when generating approximation set and scheme, threshold value is introduced into variable precision model Allowing a certain fault tolerance; Definition of candidates A kind of electronic device Dominance class The method comprises the following steps: I.e. all Not lower than The goal of solution generation is to find a minimum subset of candidates Enabling the subset to cover all good candidates with minimal subset internal conflicts; Then searching a maximum consensus cluster, and searching a maximum cluster in the conflict graph, so that no conflict edge exists between any two nodes in the cluster, wherein the maximum cluster is the candidate set with high internal consensus ; In addition, the system can dynamically adjust according to the conflict distribution of the decision, rather than using fixed And To achieve an acceptable level of consensus; Calculating the consensus degree of the scheme Calculating the definition of the scheme If the consensus and definition reach the output condition, in Internally based on global dominance Final sorting and output: Wherein, the Higher scores indicate that the candidate is better than other candidates; If the consensus is lower than the output condition, the conflict threshold is reduced And re-identify the set of conflict edges Otherwise, the dominant threshold is increased And continuing the cycle until the output condition or the threshold exceeds the preset range, and switching to the manual arbitration flow if the output condition or the threshold exceeds the preset range.
  7. 7. The method according to claim 1, wherein in step S6, the decision tree structure is a time-sequential enhancement tree structure, comprising: s6.1, each decision node contains complete time stamp information to form a decision time sequence chain; s6.2, supporting backtracking and comparison analysis of a decision process based on time; And S6.3, saving the expert experience and the high-benefit allocation scheme as special node marks for reference of subsequent decisions.
  8. 8. The method of claim 1, further comprising a risk assessment step: s7.1, performing override allocation risk detection on the generated personnel allocation scheme; S7.2, identifying a deployment mode which possibly causes instability of the team based on the historical data; And S7.3, giving an early warning prompt for the high risk scheme and suggesting an adjustment strategy.
  9. 9. A system for implementing the method of any one of claims 1-8, comprising a request receiving module, a candidate screening module, a multi-person decision interface module, a weight calculation module, a conflict resolution module, a decision tree storage module, and a risk assessment module; the request receiving module is used for acquiring personnel allocation requests; The candidate screening module is used for screening candidate persons meeting the conditions from the talent database; the multi-person decision interface module provides interfaces and functions for multiple persons to participate in decision at the same time; the weight calculation module is used for calculating the comprehensive weight coefficient of each decision maker; the conflict resolution module integrates the multiparty opinions based on the weight of the decision maker to generate an optimal scheme; the decision tree storage module is used for storing a complete decision process in a tree structure; And the risk assessment module is used for carrying out multidimensional risk assessment on the generation scheme.
  10. 10. The system of claim 9, further comprising an emotion analysis module for analyzing in real time emotion tendencies of the decision maker input text, establishing an associated model of emotion state and decision quality, and providing emotion state coefficients to the weight calculation module.

Description

Personnel allocation scheme generation method and system based on multi-person decision tracing and conflict resolution Technical Field The invention relates to the technical field of enterprise human resource management, in particular to a personnel allocation scheme generation method and system based on an artificial intelligence and a multi-person decision theory, and particularly aims at solving the problems of opinion conflict resolution and decision process traceability under a multi-person and decision scene. Background In modern enterprise management, personnel allocation is a complex and critical task, directly affecting organization efficiency and team performance. Traditional personnel allocation mainly depends on experience judgment of human resource departments, and the problems of strong subjectivity, low efficiency, lack of traceability and the like exist. With the expansion of enterprise scale and the complexity of business, personnel allocation of important posts often requires multiple department responsible persons to participate in decision making together, which makes decision making process more complex, easy to generate opinion conflict and difficult to coordinate. Some personnel deployment assistance systems have emerged in the prior art. For example, chinese patent application CN119494522a discloses an "employee post matching and adjusting method and system based on artificial intelligence", which calculates matching degree by using employee capability feature vector and post skill knowledge graph, and generates an adjusting scheme by multi-agent reinforcement learning model. The method introduces artificial intelligence technology, but mainly focuses on single person decision or algorithm automatic decision, and cannot effectively solve the opinion conflict and integration problems in multi-ginseng and decision. Another related art is chinese patent application CN118822218B, which discloses a "parameter self-optimizing person selection method" that screens persons based on condition data in person request information and calculates weights according to index data to generate recommended values. The technology considers multi-index comprehensive evaluation, but also lacks special design for a multi-person decision scene, and cannot deal with opinion divergence among decision makers. In the aspect of multi-person decision support, the prior art mostly adopts a simple voting or weighted average method, and cannot fully consider complex factors such as job level difference, emotion state, historical decision validity and the like among decision makers. In addition, most of the existing systems lack complete recording and traceability of decision processes, so that decision quality cannot be analyzed and decision modes cannot be optimized. In terms of data storage, conventional relational databases have difficulty in efficiently storing decision process data with timing characteristics and complex associations. Although time sequence database technology and tree structure storage methods are developed, a storage scheme which is optimized specifically for the characteristics of personnel allocation decision process still belongs to the blank. Therefore, the prior art has the defects that (1) a scientific multi-personnel conflict resolution mechanism is lacked, (2) the decision process is incomplete and difficult to trace back and analyze, (3) the influence of the multi-dimensional characteristics of a decision maker on the decision quality is not fully considered, and (4) the multi-dimensional risk assessment aiming at the allocation scheme is lacked in the field of enterprise personnel allocation, especially in the scene of multi-personnel and decision. The present invention aims to solve these problems. Disclosure of Invention The invention mainly aims to overcome the defects of the prior art and provide a scientific, effective, traceable and enterprise personnel allocation scheme generation method and system which are suitable for a multi-person decision scene. Specific objects include (1) The method comprises the steps of providing a set of quantitative models, scientifically integrating multiple decision opinions, effectively resolving opinion conflicts, fully recording decision making processes, forming a traceable and analyzable decision tree, supporting decision optimization, comprehensively considering multi-dimensional factors such as the job level, the emotion state, the history validity and the like of a decision maker, improving the decision scientificity, carrying out multi-dimensional risk assessment on a generated allocation scheme, pre-warning potential problems, and establishing an accumulation mechanism of expert experience and a successful scheme, and supporting continuous learning and optimization. In order to achieve the aim of the invention, the invention provides a personnel allocation scheme generation method based on multi-person decision tracing and conflict re