基于眼动技术的互联网广告效果研究*

目前，网络广告效果研究中评定广告质量的指标不够准确，广告配置科学性有待提高。因此，本文首次提出观测度的概念，将它作为衡量广告质量的新标准，并结合多种信息源对广告记忆度、兴趣度进行建模。实验结果表明，相比普遍使用显示反馈和广告自身特征，融合眼动特征来预测广告的观测程度准确性有所提高。此外，对搜索引擎结果页上位置不同，相关性不同的广告效果进行了定量分析。实验结果表明，用户的注意力受位置影响显著。广告的相关性在主体上方对注视次数影响显著，而下方和右侧表现不显著。所以可以考虑在主体下方、右侧进行与搜索内容无关商品的推广。

Research on network advertisement effect based on eye tracking technology

At present, in the study of the effect of network advertisement the measure of advertising quality not accurate enough, and the scientific nature of the ad configuration needs to be improved. So, this paper firstly propose the new concept, degree of observation, as a new criteria of the advertisement quality, and firstly combine the various information sources to model the advertising memory and interest. The experimental results show that after combining eye movement features, the accuracy to predict the effectiveness of advertising has improved. In addition, we make a quantitative analysis on the effect of the advertising in different locations and with different relevance to the main content. The experimental results show that the user''s attention is significantly affected by the location of the advertising. And the advertisement relevant to the main content has different influence on the fixation count according to its position. So, we might consider placing ads that are not related to the topic of the content to the bottom or right side of the subject.