Building a better mousetrap: Compressing mouse cursor activity for web analytics

作者：

Highlights：

• We study 10 techniques to compress users’ cursor interactions on the Web.

• A systematic evaluation of both lossy and lossless algorithms is performed.

• We found that different compression techniques excel depending on the objective.

• LZW and piecewise linear interpolation balance well accuracy and efficiency.

摘要

•We study 10 techniques to compress users’ cursor interactions on the Web.•A systematic evaluation of both lossy and lossless algorithms is performed.•We found that different compression techniques excel depending on the objective.•LZW and piecewise linear interpolation balance well accuracy and efficiency.

论文关键词：Mouse cursor tracking,Web analytics,Temporal-spatial data compression,Data simplification

论文评审过程：Received 26 December 2013, Revised 16 October 2014, Accepted 22 October 2014, Available online 15 December 2014.

论文官网地址：https://doi.org/10.1016/j.ipm.2014.10.005