Lifelog Retrieval From Daily Digital Data: Narrative Review

Area of Science:

• Lifelogging and Digital Memory Studies
• Human-Computer Interaction
• Personal Informatics

Background:

• Ubiquitous sensors in personal devices (smartphones, wearables) enable real-time data acquisition (images, audio, location, physiological signals).
• This data, termed lifelog data, facilitates analysis of personal experiences and behaviors through lifelogging.
• Decreasing sensor costs and increasing availability have driven the growth of lifelogging research.

Purpose of the Study:

• To present a narrative review of lifelogging literature from the past decade.
• To analyze lifelogging applications for information retrieval, memory support, and augmentation.
• To provide researchers with an overview of data types, methodologies, and applications in lifelogging.

Main Methods:

• Conducted a narrative review using Google Scholar and Scopus databases with lifelogging keywords.
• Screened 411 retrieved publications, fully reviewing 114 and manually including 30.
• Selected 113 publications for the review based on content analysis.

Main Results:

• Lifelogs show potential as powerful tools for memory retrieval and understanding individual experiences.
• Numerous computational tools exist for diverse lifelogging applications, utilizing multimodal data.
• Visual lifelogs are a prominent data source, with various processing approaches and algorithms identified.

Conclusions:

• Lifelogging systems offer benefits for memory augmentation and support for individuals with memory impairments.
• Lifelog data analysis can generate digital memories serving as surrogate memory.
• Contextual information extracted from lifelogs enhances understanding of daily life events, experiences, and behaviors.