Lifecycle of our research...International Conference on Research Challenges in Information Systems...

1
Intentional Process Model Discovery from Logs International Journal R. Deneckère , C. Hug, G. Khodabandelou, C. Salinesi, "Intention Mining: Process Model Discovery Using Supervised Learning", International Journal of Information System Modeling and Design (IJISMD), 2014. International Conferences G. Khodabandelou, C. Hug, C. Salinesi, "Toward an Automatic Tool for the Construction of Intentional Process Models from Event Logs", long paper, Submitted to The 8th IEEE International Conference on Research Challenges in Information Systems (RCIS), May 2014, Marrakesh, Morocco. G. Khodabandelou, C. Hug, R. Deneckère, C. Salinesi, "Supervised vs. Unsupervised Learning for Intentional Process Models Discovery", Submitted to BPMDS in conjunction with 26th International Conference on Advanced Information Systems Engineering (CAiSE), June 2014, Thessaloniki, Creek. G. Khodabandelou, C. Hug, R. Deneckère, C. Salinesi, "Unsupervised Discovery of Intentional Process Model from Event Logs", long paper, Accepted in MSR/ICSE (the 36st International Conference on Software Engineering), June 2014, Hyderabad, India. Jankovic, M., Bajec, M., G. Khodabandelou, C. Hug, R. Deneckère, C. Salinesi, "Intelligent Agile Method Framework", Proc. of the 8st International Conference on Evaluation of Novel Approaches to Software Engineering (ENASE), July 2013, Angers, France. G. Khodabandelou, C. Hug, R. Deneckère, C. Salinesi, M. Bajec, E. Kornyshova, M. Jankovic "COTS Products to Trace Method Enactment : Review and Selection “, long paper, European Conference on Information Systems (ECIS), June 2013, Utrecht, Netherlands. G. Khodabandelou, "Contextual Recommendations using Intention Mining on Process Traces", doctoral consortium, the Seventh IEEE International Conference on Research Challenges in Information Science (RCIS), May 2013, Paris, France. G. Khodabandelou, C. Hug, R. Deneckère, C. Salinesi, "Supervised Intentional Process Models Discovery using Hidden Markov Models", long paper, the Seventh IEEE International Conference on Research Challenges in Information Science (RCIS), May 2013, Paris, France. ( Best Paper Award) G. Khodabandelou, C. Hug, R. Deneckère, C. Salinesi, “Process Mining versus Intention Mining", long paper, Exploring Modelling Methods for Systems Analysis and Design (EMMSAD), June 2013, Valencia, Spain. National Conference G. Khodabandelou, C. Hug, R. Deneckère, C. Salinesi, « Découverte Supervisée des Modèles de Processus Intentionnels Utilisant Modèles de Markov Cachés ", papier long, 31th edition, National INFORSID, Mai 2013, Paris, France. Ghazaleh Khodabandelou Key Words : Intention Mining, Machine Learning, Process Traces Thesis Director : Camille Salinesi Co-directors : Rebecca Deneckère, Charlotte Hug Centre de Recherche en Informatique Context : How to discover users' intentions and strategies? Lifecycle of our research Traces Analysis _ Processes Petri nets Activities MAP Intention Intention Hypothesis: It is possible to find discover users’ intentions and strategies during process enactment out of logs. Research question: Can we discover users’ intentions and strategies during process enactment? Stakeholders Clustering Machine Learning Techniques Traces base Stakeholders Logs Map HMMs CRI- Centre de Recherche en Informatique April 2014 Activities Map discovery : Estimated Strategies Pseudo-Map Discovery : Deep Miner Algorithm Estimated Strategies Process Mining Approaches: Discovery of task-oriented model Rigid, Non-representative of humans’ rationale Intentional Model Machine Learning Users’ activities represent the real process. Map process modal allows for intentions/strategies of Information System’ users Intention Mining Discovery of the rules among logs: Strategies BPMN Intention Traces base Supervised Learning Unsupervised Learning Map Miner Algorithm Specify an entity Specify an association Sub-Intentions Specify an entity Specify an association Sub-Intentions Pseudo-Map Pseudo-Map Discovered Map Process Model

Transcript of Lifecycle of our research...International Conference on Research Challenges in Information Systems...

Page 1: Lifecycle of our research...International Conference on Research Challenges in Information Systems (RCIS), May 2014, Marrakesh, Morocco. G. Khodabandelou, C. Hug, R. Deneckère, C.

Intentional Process Model Discovery from Logs

International Journal R. Deneckère , C. Hug, G. Khodabandelou, C. Salinesi, "Intention Mining: Process Model Discovery Using Supervised Learning", International Journal of Information System Modeling and Design (IJISMD), 2014. International Conferences G. Khodabandelou, C. Hug, C. Salinesi, "Toward an Automatic Tool for the Construction of Intentional Process Models from Event Logs", long paper, Submitted to The 8th IEEE International Conference on Research Challenges in Information Systems (RCIS), May 2014, Marrakesh, Morocco. G. Khodabandelou, C. Hug, R. Deneckère, C. Salinesi, "Supervised vs. Unsupervised Learning for Intentional Process Models Discovery", Submitted to BPMDS in conjunction with 26th International Conference on Advanced Information Systems Engineering (CAiSE), June 2014, Thessaloniki, Creek. G. Khodabandelou, C. Hug, R. Deneckère, C. Salinesi, "Unsupervised Discovery of Intentional Process Model from Event Logs", long paper, Accepted in MSR/ICSE (the 36st International Conference on Software Engineering), June 2014, Hyderabad, India. Jankovic, M., Bajec, M., G. Khodabandelou, C. Hug, R. Deneckère, C. Salinesi, "Intelligent Agile Method Framework", Proc. of the 8st International Conference on Evaluation of Novel Approaches to Software Engineering (ENASE), July 2013, Angers, France. G. Khodabandelou, C. Hug, R. Deneckère, C. Salinesi, M. Bajec, E. Kornyshova, M. Jankovic "COTS Products to Trace Method Enactment : Review and Selection “, long paper, European Conference on Information Systems (ECIS), June 2013, Utrecht, Netherlands. G. Khodabandelou, "Contextual Recommendations using Intention Mining on Process Traces", doctoral consortium, the Seventh IEEE International Conference on Research Challenges in Information Science (RCIS), May 2013, Paris, France. G. Khodabandelou, C. Hug, R. Deneckère, C. Salinesi, "Supervised Intentional Process Models Discovery using Hidden Markov Models", long paper, the Seventh IEEE International

Conference on Research Challenges in Information Science (RCIS), May 2013, Paris, France. (Best Paper Award) G. Khodabandelou, C. Hug, R. Deneckère, C. Salinesi, “Process Mining versus Intention Mining", long paper, Exploring Modelling Methods for Systems Analysis and Design (EMMSAD), June 2013, Valencia, Spain. National Conference G. Khodabandelou, C. Hug, R. Deneckère, C. Salinesi, « Découverte Supervisée des Modèles de Processus Intentionnels Utilisant Modèles de Markov Cachés ", papier long, 31th edition, National INFORSID, Mai 2013, Paris, France.

Ghazaleh Khodabandelou Key Words : Intention Mining, Machine Learning, Process Traces

Thesis Director : Camille Salinesi Co-directors : Rebecca Deneckère, Charlotte Hug Centre de Recherche en Informatique

Context : How to discover users' intentions and strategies?

Lifecycle of our research

Traces Analysis

_

Processes

Petri nets

Activities

MAP

Intention

Intention

Hypothesis: It is possible to find discover users’ intentions and strategies during process enactment out of logs.

Research question: Can we discover users’ intentions and strategies during process enactment?

Stakeholders

Clustering

Machine Learning

Techniques

Traces base

Stakeholders Logs

Map

HMMs

CRI- Centre de Recherche en Informatique April 2014

Activities

Map discovery :

Estimated Strategies

Pseudo-Map Discovery :

Deep Miner Algorithm

Estimated Strategies

Process Mining Approaches: Discovery of task-oriented model Rigid, Non-representative of humans’ rationale

Intentional Model Machine Learning

Users’ activities represent the real process.

Map process modal allows for intentions/strategies of Information System’ users

Inte

nti

on

Min

ing

Discovery of the rules among logs: Strategies

BPMN

Intention

Traces base

Supervised Learning

Unsupervised Learning

Map Miner Algorithm

Specify an entity Specify an association

Sub-Intentions

Specify an entity Specify an association

Sub-Intentions

Pseudo-Map

Pseudo-Map Discovered Map Process Model