MapReduce-guided scalable compressed dictionary construction for evolving repetitive sequence streams

Abstract

Users' repetitive daily or weekly activities may constitute user profiles. For example, a user's frequent command sequences may represent normative pattern of that user. To find normative patterns over dynamic data streams of unbounded length is challenging. For this, an unsupervised learning approach is proposed in our prior work by exploiting a compressed/quantized dictionary to model common behavior sequences. This work suffers scalability issues. Hence, in this paper, we propose and implement a MapReduce-based framework to construct a quantized dictionary. We show effectiveness of our distributed parallel solution on a benchmark dataset.