【摘要】 收视率是衡量电视节目质量的重要指标,而质量不高的收视调查数据将导致收视率指标计算的不准确。为解决基于数据挖掘的电视收视率分析的数据问题,就调查数据进行了有效的前期处理,其中涉及原始数据的正确录入,对存入数据库的原始数据实施的数据清理、数据变换、数据归约技术和应用的模糊集理论。实验结果表明在预处理后的收视率资料库中能够挖掘出知识,为收视率应用提供了有力支持。更多还原

【Abstract】 Audience rating is an important parameter that scales TV programs’ quality,but rating surveying data of low quality lead to much inaccurate on calculating audience rating.To solve the data problem of TV audience rating based on data mining,the preprocessing technologies of surveying data are discussed,which involve the sorting out of original data,data cleaning,data transformation、data reduction technologies and fuzzy set theory.Experimental results indicate that knowledge from preprocessed audience rating database be can gotten,provided powerful support for applying audience rating.更多还原