Cued Retrieval of Personal Memories of Social Interactions

Abstract

This paper aims at developing a social interactions summarizer system that firstly summarizes a person's daily social interactions, with the purpose of enhancing his episodic memory and secondly provides various methods for searching in the collected data. The first goal originates from studies that have shown that replaying video or audio recordings of an experience have proved effective in enhancing people's episodic memory. The second goal is based on the fact that with the emergence of wearable devices for lifelogging, every day huge archives of data consisting of images, audio, etc could be generated from a person's life. Over time the growth rate of such archives is so substantially high that manually searching them, even after a few months of data collection, would be cumbersome and virtually impossible. In this study, we present a method for effectively summarizing one's social interactions for enhancing her episodic memory. Our results, reporting work in progress, illustrate that our method is highly effective in augmenting one's episodic memory.