Self-supervised mining of human activity from CGM

Abstract

The goal of this paper is to describe a method to automatically extract all basic attributes namely actor, action, object, time and location which belong to an activity, and the transition between activities in each sentence retrieved from Japanese CGM (consumer generated media). Previous work had some limitations, such as high setup cost, inability of extracting all attributes, limitation on the types of sentences that can be handled, and insufficient consideration of interdependency among attributes. To resolve these problems, this paper proposes a novel approach that treats the activity extraction as a sequence labeling problem, and automatically makes its own training data. This approach has advantages such as domain-independence, scalability, and unnecessary hand-tagged data. Since it is unnecessary to fix the positions and the number of the attributes in activity sentences, this approach can extract all attributes and transitions between activities by making only a single pass over its corpus. © 2010 Springer-Verlag Berlin Heidelberg.