Mining frequent items in the time fading model

TL;DR

FDCMSS is a novel sketch-based algorithm for efficiently mining frequent items in data streams under the time fading model, combining decay, Count-Min, and Space Saving techniques.

Contribution

It introduces a new algorithm that outperforms existing methods in speed, space efficiency, and accuracy for time-fading data stream mining.

Findings

FDCMSS outperforms λ-HCount in experiments.

The algorithm is faster and uses less space.

It achieves higher precision and lower error.

Abstract

We present FDCMSS, a new sketch-based algorithm for mining frequent items in data streams. The algorithm cleverly combines key ideas borrowed from forward decay, the Count-Min and the Space Saving algorithms. It works in the time fading model, mining data streams according to the cash register model. We formally prove its correctness and show, through extensive experimental results, that our algorithm outperforms $\lambda$-HCount, a recently developed algorithm, with regard to speed, space used, precision attained and error committed on both synthetic and real datasets.