Mining Users' Intents from Logs

Mining Users' Intents from Logs

Ghazaleh Khodabandelou (Centre de Recherche en Informatique, Université Paris 1 Panthéon – Sorbonne, Paris, France & Department of Computer Science, University of Paderborn, Paderborn, Germany), Charlotte Hug (Centre de Recherche en Informatique, Université Paris 1 Panthéon – Sorbonne, Paris, France) and Camille Salinesi (Centre de Recherche en Informatique, Université Paris 1 Panthéon – Sorbonne, Paris, France)
Copyright: © 2015 |Pages: 29
DOI: 10.4018/IJISMD.2015040102

Abstract

Intentions play a key role in information systems engineering. Research on process modeling has highlighted that specifying intentions can expressly mitigate many problems encountered in process modeling as lack of flexibility or adaptation. Process mining approaches mine processes in terms of tasks and branching. To identify and formalize intentions from event logs, this work presents a novel approach of process mining, called Map Miner Method (MMM). This method automates the construction of intentional process models from event logs. First, MMM estimates users' strategies (i.e., the different ways to fulfill the intentions) in terms of their activities. These estimated strategies are then used to infer users' intentions at different levels of abstraction using two tailored algorithms. MMM constructs intentional process models with respect to the Map metamodel formalism. MMM is applied on a real-world dataset, i.e. event logs of developers of Eclipse UDC (Usage Data Collector). The resulting Map process model provides a precious understanding of the processes followed by the developers, and also provide feedback on the effectiveness and demonstrate scalability of MMM.
Article Preview

Introduction

Process mining approaches aim at modeling users' behaviors in terms of sequences of tasks and branching in an automatic way (Van der Aalst, 2004) (Agrawal, 1998) (Cook, 1998) (Datta, 1998) (van Dongen, 2004) (Weijters, 2003) (Herbst, 2000) where the mined process models are activity-oriented models. However, processes can be seen as teleological (Ralph, 2008). A teleological process is a process that takes into account the teleological behaviors of process enactment (behaviors attached to the notion of goal). It describes the intentions (goals, objectives) associated with a result that an individual wants to obtain. In the late 90s, Rolland introduced a new category of process models, called intentional process models (Rolland, 1999), which takes into account the notions of intention and strategy to model the process enactment. A strategy is an approach, a manner or a means to achieve an intention (Rolland, 2007). Specifying intentions and strategies has proved to be a powerful tool to better understand the deep nature of processes, to see how processes interweave and combine, to abstract processes and visualize them under man-manageable form, even when they are extremely complex (Rolland, 2005). Intentional process models have emerged to offer a flexible structure to model processes. Many research works in intentional process modeling demonstrate that the fundamental nature of processes is mostly intentional and the processes should be modeled from an intentional point of view (Davis, 1989) (Plihon, 1996) (Rolland, 1999). According to these approaches, an enacted process is a reflection of humans' intention performed as a sequence of activities. Therefore, it is not possible to model humans' cognitive operators, e.g., thinking, deciding, and acting process only in terms of a simple sequence of activities. Indeed, an intention is a goal that a user wants to achieve regarding the context in which he/she is working (Plihon, 1996). The notion of context plays a key role for the intention, since a given intention emerges in a given context, which not only promotes its appearance, but also influences the realization of this intention (Rolland, 2005). In the method engineering context (Jankovic, 2013), it is essential to capture intentions that led to the implementation of activities to understand methods used by stakeholders and their ways of working.

Intentions are a first class concept of information systems engineering (Rolland, 2005). In the early 80s, intentional process models have been proposed in information systems community (Swanson, 1974) (Christie, 1981) as a potential theoretical foundation to determine user's behavior (Davis, 1989). Intention models take root in a former work Technology Acceptance Model (TAM) (Davis, 1989) one of the extensions of Theory of Reasoned Action (TRA) (Ajzen, 1975) designed to model humans' behavioral intention, especially for computer usage behavior.

The TRA has proven effective in predicting and explaining humans' behavior through various domains. Since the early 90s, intentional software process specification have been promoted as a driving paradigm to study strategic alignment (Thévenet, 2007) (Etien, 2006) to define actors and roles, to specify the outcome of business process models (Salinesi, 2003) and name them, to analyze, to support guidance (Rolland, 1993) (Deneckère, 2010) to describe intentional services (Rolland, 2010), to handle traceability issues (Jarke, 1993) to express pervasive information systems (Najar, 2011) to define systems requirements (Ralyté, 1999), to study users' behavior to identify and name use cases, to tailor methods (Ralyté, 2003) or to design more flexible methods (Mirbel, 2006), etc. Further, research on guidance in method engineering shows that many method engineering issues, such as rigidity or lack of adaptation, are solved more effectively when intentions and strategies are explicitly specified (Rolland, 2005).

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 10: 4 Issues (2019): Forthcoming, Available for Pre-Order
Volume 9: 4 Issues (2018): 3 Released, 1 Forthcoming
Volume 8: 4 Issues (2017)
Volume 7: 4 Issues (2016)
Volume 6: 4 Issues (2015)
Volume 5: 4 Issues (2014)
Volume 4: 4 Issues (2013)
Volume 3: 4 Issues (2012)
Volume 2: 4 Issues (2011)
Volume 1: 4 Issues (2010)
View Complete Journal Contents Listing