Research On Fine-grained Sentence Prediction Technology Based On Sentencing Principle

Since the 18th NATIONAL Congress of the CPC,the new Party Central Committee with Comrade Xi Jinping at its core has carried out comprehensive and in-depth judicial reform,breaking new ground in building a country under the rule of law.However,at the current stage,there are still problems of "isolation" and "poverty" in the administration of justice.The wide application of AI provides a new idea to solve the above problems.Intelligent justice,as one of the hottest keywords at present,is in the stage of rapid development.The sentence prediction task is an important research topic of Intelligent justice.In order to further improve judicial fairness,starting from the classification and prediction of sentence distribution,aiming at the problems of actual criminal cases,this paper puts forward a new method of word segmentation feature screening through data set acquisition,data cleaning,text word segmentation,positive sampling and model fusion.On the basis of removing stop words and ground nouns,this word segmentation method further extracts all word segmentation including keywords through the new screening method proposed in this paper.After the screening of word segmentation features,the over sampling method is introduced to reduce the negative impact of over fitting on the classification results.Combined with the voting mechanism and model fusion method,it is applied to the practice of the classification and prediction task of sentence distribution,which effectively improves the classification and prediction accuracy of sentence distribution.Through a series of research experiments,this paper proposes a new segmentation feature screening method,and introduces the over sampling method,combined with the voting mechanism and model fusion method,which is applied to the classification and prediction of the distribution of prison terms,which effectively improves the classification accuracy.The research work carried out in this paper not only provides a new idea for the development of the field of Intelligent justice,especially in the field of classification and prediction of sentence distribution,but also contributes to the realization of judicial fairness and justice.