Retrieval Mechanism Of Similar Cases And Related Cases

It has been more than two years since the similar case retrieval mechanism was put forward on August 1,2017.The Supreme People’s court and some local courts have intensively issued Implementation norms,and there have been a lot of similar case retrieval practices in the judgment.It is urgent for us to review and reflect on the design and implementation of this mechanism.This paper attempts to complete this task.From the theoretical and practical level,this paper combs the operation of the case retrieval mechanism in judicial practice,and on this basis,from the perspective of judicial principles and judicial practice,puts forward some countermeasures to improve the case retrieval mechanism,in order to make the case retrieval mechanism better realize its goal of standardizing the discretion of judges and unifying the judgment scale.In addition to the introduction and conclusion,the paper is divided into five parts,totaling more than 20000 wordsThe first part is about the emergence and development of case retrieval mechanism.Similar case retrieval mechanism comes into being in the context of the reform of judicial responsibility system,supported by big data and artificial intelligence technology.Since the emergence of case retrieval mechanism,relevant departments have issued a large number of implementation norms.This part combs these implementation norms,clarifies the development context of the case retrieval mechanism,in order to form a clearer understanding of the case retrieval mechanism.The second part is the basic content and working principle of the case retrieval mechanism.The basic contents of the case retrieval mechanism are: the judges rely on the case retrieval system to comprehensively retrieve the cases similar to the cases to be handled,the judges make the case retrieval reports based on the retrieval results,the judges compare the judgment scale of the same effective cases with the cases under trial,and make the judgment based on the case classification.The working principle of class case retrieval mechanism is mainly reflected in the operation of class case retrieval system.First of all,case retrieval system relies on knowledge map technology and natural semantic processing technology to structure unstructured and semi-structured judicial big data.Secondly,artificial intelligence identifies these structured judicial big data and conducts in-depth learning,so as to obtain a high confidence algorithm.Based on the high confidence algorithm,the case retrieval system can retrieve and push the case.The third part,the implementation effect of case retrieval mechanism.The implementation of similar case retrieval mechanism provides judges with reference to the judgment scale of similar cases,prevents judges from making decisions that are too different from the judgment scale of similar cases,and is conducive to unifying the judgment scale of various cases.Secondly,the implementation of case retrieval mechanism improves the efficiency of handling cases,and the improvement of efficiency makes judges focus more on improving the quality of cases.Finally,the judge can cite the results of case retrieval as the reason of judgment in the judgment document,strengthen the reasoning of the judgment document,so as to enhance the credibility of the judgment.The fourth part,the implementation dilemma of case retrieval mechanism.At present,the case retrieval mechanism still faces many difficulties in practice.The biggest difficulty is that the case retrieval system does not accurately retrieve and push class cases.In fact,some class cases do not meet the standard of class cases,and some class cases,although in similar form,have no reference value.The second dilemma of the retrieval mechanism is the imbalance of the number of cases pushed by retrieval.The quantity imbalance is manifested in that sometimes the system pushes too many class cases,while sometimes it pushes too few class cases.Too many cases will increase the difficulty for judges to find suitable cases,and too few cases will not be enough to form the judgment scale of such cases.The third implementation dilemma is that the retrieval system of similar cases pushed by similar cases is chaotic and lack of order.The specific performance is in the retrieval of pushed cases,the scope is relatively centralized,the source is unknown,the level is unclear.The fifth part is about the causes of the dilemma of case retrieval mechanism.There are many reasons for the dilemma of similar case retrieval mechanism.The primary reason is that the data of referee documents are not comprehensive and the data quality of some referee documents is not good.Secondly,the current big data and artificial intelligence technology are not mature,which leads to "similar but useless" cases pushed by retrieval.Third,the classification of classification standards is not reasonable,and some system searches and pushes class cases can not meet the standards of class cases.Finally,the imperfection of retrieval rules deepens the dilemma of case retrieval mechanism in practice.The imperfection of retrieval rules is mainly reflected in the following aspects: the lack of specific operating procedures and unified case retrieval system,the failure to clearly stipulate the effectiveness of reference cases,the failure to consider the impact of judicial practice and changes in legal provisions on the effectiveness of reference cases.The sixth part is the countermeasures to improve the retrieval mechanism of similar cases.On the basis of the first two parts,starting with the essential attribute and function source of the case retrieval mechanism,this part puts forward the countermeasures to improve the case retrieval mechanism.First of all,we need to optimize a large amount of unordered judgment document data into comprehensive,orderly and high-quality data.Secondly,we should give full play to the subjective initiative,improve the technical application means,and improve the case retrieval system to the greatest extent.Third,we should draw lessons from the factor based trial law and take the type factor as the standard to identify the cases.Finally,improve the retrieval rules: formulate the national unified retrieval operation rules and retrieval system of similar cases,clearly stipulate the effectiveness of reference cases,correctly evaluate and deal with the effectiveness of changed cases.