Graph clustering based size varying rules for lifelong topic modelling

Abstract

Lifelong learning topic models identify the hidden concepts discussed in the collection of documents. Lifelong learning models have an automatic learning mechanism. In the learning process, the model gets more knowledgeable with experience as it learns from the past in the form of rules. It carries rules to the future and utilises them when a similar scenario arise in the future. The existing lifelong learning topic models heavily rely on statistical measures to learn rules that lead to two limitations. In this research work, we introduce complex networks analysis for learning rules. The rules are obtained through hierarchical clustering of the complex network that has different number of words within a rule and has directed orientation. The proposed approach improves the utilisation of rules for improved quality of topics at higher performance with unidirectional rules on the standard lifelong learning dataset.

Keywords: networks, lifelong, models, networks analysis