基于眼动数据的用户在线购物时间压力识别研究

Research on Time Pressure Recognition of Online Shopping Users Based on Eye Movement Data

【目的/意义】旨在基于眼动数据建立机器学习模型以识别出用户购物时的时间压力水平。【方法/过程】共招募了32名被试进行了一项有关购物的眼动追踪实验,让被试在不同的时间压力水平下执行四项任务,选择随机森林、支持向量机、梯度提升树和k近邻等机器学习算法构建识别模型,利用准确率、查全率、查准率、F1值和ROC等指标评估模型。【结果/结论】随机森林有着最好的识别精度,在测试集上的预测准确率达到了87.5%,其中注视持续时间和注视次数等注视类眼动指标为识别模型贡献最大。

[Purpose/significance]This paper aims to establish a machine learning model based on eye movement data to identify the time pressure level of users when shopping.[Method/process]The paper recruits a total of 32 subjects to conduct an eye tracking experiment on shopping.The subjects perform four tasks under different levels of time pressure,select machine learning algorithms such as random forest,support vector machine,gradient boosting tree and k-nearest neighbor to construct the recognition model,use the accuracy rate,recall rate,precision rate,the values of F1 and ROC index assessment model.[Result/conclusion]Random forest has the best recognition accuracy,and the prediction accuracy on the test set reaches 87.5%,in which the fixation eye movement indicators such as fixation duration and fixation number contribute the most to the recognition model.