Research On Strategy-Proof Voting And Improvement Of Preference Aggregation Algorithm

The voting problem is closely related to our life and work,from political elections to various questionnaires,which plays a positive role in promoting social democracy,but there are still some problems to be solved.This paper will study and discuss the voting problem from the aspects of the computational complexity of manipulation,the strategy-proof and the improvement of preference aggregation algorithm.First,this paper expounds the importance of the research on the strategy-proof voting rules,introduces the foundation of social choice and voting theory,and lists several common voting rules:position scoring rule,Condorcet expansion rule and other rules.Then,the theory foundation of strategy-proof voting,Impossibility Theorem,is introduced briefly.The computational complexity of different manipulation behaviors under several voting rules is studied,and the better voting rules are found out by analyzing and comparing them.Then,the author compares several methods of Kemeny ranking aggregation,which has good strategy-proofness.Our aim is to establish the best balance between search time and performance.For the NP-hard problem in theory,we find that the problems in practice span the three states:strong consensus,weak consensus and no consensus.We also give specific suggestions for each and propose a fast test to distinguish different situations.Although there are various algorithms,few of them are always Pareto optimal.Finally,the paper studies the optimization of the aggregation of weighted partial ordering based on Kemeny rule by using the improved ant colony algorithm.Preference aggregation is a kind of problem in the field of social computing,which has received a lot of attention and has been widely used in practice.It studies how to aggregate and integrate the preferences of individuals in a group to get the unified preferences of the whole group.When a group wants to make a decision,but the preferences of each person in the group may be different.At this time,we need to use the method of preference fusion to get a unified preference,so as to make the corresponding decision on behalf of the whole group.In many cases,due to the number of alternatives,the voters may not give a complete ranking.How to aggregate the ranking in line with the consensus of the group is our research focus.In a word,this paper discusses and studies several common voting rules in detail,as well as the computational complexity of strategy voting under different rules.It has a more in-depth understanding of strategy-proof voting,and studies and improves the preference aggregation algorithm.The results prove that it has better properties.