排序支持的交互数据分类算法及其应用

交互分类是解决数据复杂分类问题的主要手段之一。在现有的大多交互分类系统中，用户能准确识别数据类别，但在有些分类场景中,类别之间的顺序关系更容易被识别，为此，提出一种排序支持的交互数据分类算法。为提升交互分类精度，引入数据的顺序信息，为降低标记难度，提出候选样本推荐策略。另外,提出一种评估分类算法性能的可视化方法，用包含基本车况、交通违法记录、交通事故记录等信息的车辆数据集进行实验验证，将相关车辆分为高危车辆、中危车辆、低危车辆3类，算法的分类结果模型一致度达近98%，验证了方法的有效性。

Ranking-supported interactive data classification method and its application

Interactive classification is one of the main means schemes to solve complex classification problems of data.In most existing interactive classification system,users can accurately identify the categories of data.However,in some classification scenarios,there is an order relationship between categories,and it is easier for users to identify the order relationship. This paper presented an interactive data classification method supported by sorting.It introduces the user's cognition of the data order information to improve the accuracy of interactive classification with a candidate sample recommendation strategy to reduce users' difficulty of identifying.To demonstrate the potential of the new approach, we propose a visualization method for classification algorithm performance measurement based on the vehicle data set of accident information including basic vehicle information,illegal driving information and vehicle traffic records.The relevant vehicles are divided into high-risk vehicles,medium-risk vehicles,and low-risk vehicles.The classification consistency reaches nearly 98%,which verifies the effectiveness of the proposed method.