NTCIR Lifelog

Abstract

Test collections have a long history of supporting repeatable and comparable evaluation in Information Retrieval (IR). However, thus far, no shared test collection exists for IR systems that are designed to index and retrieve multimodal lifelog data. In this paper we introduce the first test collection for personal lifelog data, which has been employed for the NTCIR12-Lifelog task. In this paper, the requirements for the test collection are motivated, the process of creating the test collection is described, along with an overview of the test collection. Finally suggestions are given for possible applications of the test collection.

Building a Disclosed Lifelog Dataset

“…Learning from the NTCIR-12 [6,7] and the LTA-2016, we propose these principles for building a disclosed lifelog dataset:…”

“…However, the design and construction of a lifelog dataset that will be shared and studied in public is not trivial, where there are signi cant technical challenges to be solved, arising from the gathering, semantic enrichment, and pervasive accessing of these vast personal data archives [6]. To the best of our knowledge, there has been only a published lifelog dataset that ful lled major of the problems above: the NTCIR-12 -Lifelog [6,7].…”

“…e goal of a published dataset is to share with researchers. Lifelogs, however, di er from more traditional shared datasets [6], e.g., the privacy-aware issues, and thus need a well designed strategy to control data access and modify (if needed) the lifelog data.…”

“…In this study, we aim at proposing principles for construction of a disclosed lifelog dataset as well as describe the process of how we built a new lifelog dataset, the NTCIR-13 -Lifelog 2 6 , following the proposed principles. e contributions of this paper are: (i) we point out the challenges for building a shared lifelog dataset; (ii) we propose principles for building a shared lifelog dataset (Section 2); (iii) based on the proposed principles, we build and describe the whole processes from data gathering to determine the roles for the people who are building, sharing and exploiting a disclosed lifelog dataset (Section 3); (iv) a framework for anonymisation (Section 4); and (v) a short list of potential applications are recommended (Section 5).…”

Lifer

“…One of the interesting use cases is data retrieval and accessibility applications; allowing the individual to search and access previous life events in e cient and timely manner to behave as an external digital memory that can be called upon for di erent purposes. is is yet unsolved challenge and to address this, collaborative benchmarking exercises and workshops have been organised recently such as NTCIR-13 -Lifelog [5] and ImageCLEF2017lifelog [10]. We note that there is still no standard or typical approach to developing a lifelog search engine arising from these benchmarking exercises.…”