On integrating heterogeneous lifelog services

This paper presents a framework that integrates different kinds of lifelog services. For efficient data mashup, we first propose the lifelog common data model (LLCDM), which normalizes data structures and formats of heterogeneous lifelog records. We derive application-neutral data items by an interrogative analysis of what, why, when, who, where and how. We then implement the lifelog mashup API (LLAPI) to achieve standardized access to heterogeneous lifelogs. A case study of integrating practical lifelog services (Twitter, Flickr and GARMIN Connect) demonstrates the effectiveness of the proposed framework. It was shown that the development effort with the proposed APIs was reduced to 11.9%, compared to the conventional mashup development with the proprietary APIs.