数字电视收视率预测模块设计与实现

发布时间：2018-02-26 06:00

本文关键词： 数字电视收视率模块设计预测网络　出处：《电子科技大学》2010年硕士论文　论文类型：学位论文

【摘要】：电视频道收视率是一项非常重要的指标,不仅关系到电视台频道在群众中的口碑并且对电视台自身经营效益也有巨大的影响;另一方面需要在电视台购买广告时间的企业客户而言,电视台收视率的高低也直接影响到购买广告时间价格的高低,对那些希望更好的宣传自己产品的企业来说高的电视台收视率无疑会带来更好的效果没有竞争力的频道则无疑只会浪费企业的时间与金钱,或者企业会选择将一些不重要的品牌放到这些具有低收视率的频道播放。在本文中主要研究了电视的收视率数据的测试与预测模块,主要是从特定的时间段的角度进行了论证和分析研究,电视收视率测试数据是以时间为序列来进行表述的,是将每天的某个预定的时间段来进行提取和分析电视的收视率,在此时间周期段内是按照分钟数每间隔一分钟对收视率进行收集这样一来就可以得到数量巨大的测试数据,采用是基于时间的序列来进行分析并以分类预测的算法对电视的收视率进行科学的预测。本软件系统预测模块主要功能包括如下的几个部分:一是系统模块的逻辑与控制部分,其收视率模块来负责控制整个系统逻辑运算与控制;二是对数据进行输入的功能模块,这部分提供对各类的格式的数据的输入并对测试数据进行相应的预处理并将读入的数据以某种特定的方式进行相应的运算并存储,这里面涉及到各类数据结构其目的是如何去满足各种不同的应用。本研究课题首先论述了电视收视率预测的国内外现状随后后提出了收视率数据形式且对于收视率得测试数据,本研究课题的创新在于利用数据挖掘的分类技术预测频道收视率的方法,包括对贝叶斯分类及决策树的分类与学习,并且对于实际的测试数据而建立了模型进行了相关的预测分析。使用数据的挖掘分类的方法在理论上是行得通的但只能预测不知道的收视率测试数据并且不能得到实际的收视率数据值,本研究课题提出了相应的软件系统,并对该系统主要框架、主要模块及相互间的交互行为、主要接口进行了设计来解决频道的收视率的预测的问题,该系统模块就目前的需求与和潜在的不知道的需求领域按照软件工程的基本原则利用面向对象的思想进行的设计来尽量的减小模块之间的耦合性使系统具有很好的可扩展性并留有将来的升级接口。
[Abstract]:TV channel rating is a very important indicator, which not only relates to the public reputation of TV channel, but also has a huge impact on the operating efficiency of TV station itself. On the other hand, for corporate customers who need to buy advertising time in television stations, the ratings of television stations also directly affect the price of the time they buy advertisements. For companies that want to promote their products better, high ratings will no doubt lead to better results. Uncompetitive channels will no doubt only waste their time and money. Or companies will choose to put unimportant brands on channels with low ratings. In this paper, we mainly study the test and prediction module of TV ratings data, mainly from the point of view of a specific time period. The test data of TV ratings are expressed in time series. Is to extract and analyze the TV ratings at a predetermined time period of the day, and in this period of time, to collect the ratings per minute of minutes, so that you can get a huge amount of test data. This software system forecast module mainly includes the following several parts: first, the logic and control part of the system module, the first part is the logic and control part of the system module, and the main functions of the software system prediction module are as follows: first, the logic and control part of the system module, the main functions of the software system prediction module are as follows: 1. The ratings module is responsible for controlling the logic operation and control of the whole system. This section provides the input of data in various formats and the corresponding preprocessing of the test data and the operation and storage of the read data in a certain way. This involves all kinds of data structures whose purpose is to meet a variety of different applications. This research topic first discusses the domestic and foreign present situation of TV ratings prediction and then puts forward the form of ratings data and gets the test data for ratings. The innovation of this research is to use the classification technology of data mining to predict channel ratings, including the classification and learning of Bayesian classification and decision tree. The method of mining and classifying data using data is feasible in theory but can only predict the unknown ratings test data and can not be obtained. To the actual ratings data value, In this paper, the corresponding software system is put forward, and the main frame, the main modules and the interaction between each other, the main interface are designed to solve the problem of channel rating prediction. According to the basic principles of software engineering, the system module is designed to minimize the coupling between modules according to the basic principles of software engineering to minimize the coupling between modules. Extensibility with future upgrade interfaces.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2010
【分类号】：TN949.197;G223

【参考文献】