Mining goal refinement patterns: Distilling know-how from data

Authors

Metta Santiputri, University of WollongongFollow
Novarun Deb, University of Calcutta
Muhammad Khan, University of WollongongFollow
Aditya K. Ghose, University of WollongongFollow
Hoa K. Dam, University of WollongongFollow
Nabendu Chaki, University of Calcutta

RIS ID

117346

Publication Details

Santiputri, M., Deb, N., Khan, M., Ghose, A., Dam, H. & Chaki, N. (2017). Mining goal refinement patterns: Distilling know-how from data. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (pp. 69-76). Switzerland: Springer.

Abstract

Goal models play an important role by providing a hierarchic representation of stakeholder intent, and by providing a representation of lower-level subgoals that must be achieved to enable the achievement of higher-level goals. A goal model can be viewed as a composition of a number of goal refinement patterns that relate parent goals to subgoals. In this paper, we offer a means for mining these patterns from enterprise event logs and a technique to leverage vector representations of words and phrases to compose these patterns to obtain complete goal models. The resulting machinery can be quiote powerful in its ability to mine know-how or constitutive norms. We offer an empirical evaluation using both real-life and synthetic datasets.