BIGFile

Bayesian Information Gain for Fast File Retrieval

BIGFile is a fast navigation-based file retrieval technique where the computer is trying to gain information from the user by providing shortcuts that may help access the target faster. These shortcuts are presented in a split adaptive area of a file retrieval interface and include the estimated files or folders selected by our computationally efficient algorithm BIGFileFast, which together with the items in the current folder, maximize the expected information gain from the next user input. The interface includes the paths to the estimated items so that contextual information is provided to identify them. Users can use any shortcut in the adaptive area or simply navigate the hierarchy as usual.

Results of an experiment comparing BIGFile with a Finder-like list view show that BIGFile was up to 64% faster than Finder, and users unanimously preferred the split interfaces.

BIGFile is the second example of the framework Bayesian Information Gain (BIG) that I developped during my Ph.D. thesis.

An overview of BIGFile is here:

Publication

Wanyu Liu, Olivier Rioul, Joanna McGrenere, Wendy E. Mackay, and Michel Beaudouin-Lafon. 2018. BIGFile: Bayesian Information Gain for Fast File Retrieval. In Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems (CHI '18). Association for Computing Machinery, New York, NY, USA, Paper 385, 1–13. DOI:https://doi.org/10.1145/3173574.3173959