项目名称: 配音演员的声音对广告效果的影响--基于机器学习的声音广告研究
项目编号: No.71472192
项目类型: 面上项目
立项/批准年度: 2015
项目学科: 管理科学
项目作者: 王毅
作者单位: 中央财经大学
项目金额: 62万元
中文摘要: 随着汽车的快速普及和移动电子设备的兴起,以广播广告为代表的声音广告近些年来获得飞速成长。在声音广告中,配音演员的声音是广告的主角,但是关于配音演员的声音如何对广告效果产生影响的研究目前还相对较少。 有鉴于此,本项目将在前期研究的基础上,从计算机科学领域借鉴相关的机器学习等大数据研究工具,通过多种渠道获取、存储并建立相关的声音广告数据库,分析、提取并构建广告声音特征模型,对各项声音特征进行广告效果测量,在此基础上形成广告任务-声音匹配模型,并深入探索声音影响听众感知的内在机制,形成声音-意想-态度链理论模型。 本项目属于跨领域的交叉学科研究,预期的研究成果对于广告界来说具有很强的应用性和延伸性,可以帮助企业为广告挑选出适合的声音并建立起定量、规范的广告配音演员选择体系。
中文关键词: 声音广告;机器学习;声音特征;广告效果
英文摘要: With the rapid popularizing of automobile and mobile electronic devices, the audio advertisement got very fast progress in recent years, in which the radio advertisement was the outstanding representative. In audio advertisement, the voice of dubber is the protagonist. But there are rare researches on the impact of dubber's voice on advertising performance. This project will focus on this question using the big data research method from computer science. We will collect, store, and analyze the audio data from multi-channel, and form the audio database for the research. We will extract and analyze the audio features and construct the advertisement voice features model, advertisement task-voice matching model and voice-imagine-attitude chain. This project is an interdisciplinary research plan, the expected result will be very useful and expansible for the advertisers, they will help the advertiser to find out the most proper voice for the advertisement task and help them to select and training the dubbers.
英文关键词: audio advertisement;machine learning;audio feature;advertisement performance