The work within the project is conducted within the Polish National Science Center research grant for young researches without the Ph.D. degree, project "Recommendation method RMV for partner and service selection in virtual organization breeding environments based on process mining techniques".

In this project a novel method for supporting partner and service selection processes based on process mining techniques is proposed. The challenge and scientific goal is to analyze past instances of collaborative processes of partner and service selection, and use this knowledge to support current and future instances. The method allows activity patterns in various instances of partner and service selection processes to be identified as repeatable patterns occurring in particular contexts. Such patterns can be used for recommendation of activities. This provides an opportunity for better computer support of collaborative processes (e.g. by automation of activity patterns) leading to more efficient and effective realization of partner and service selection process goals. Within the project, efficient algorithms will be designed and implemented, namely: activity pattern identification, context modeling and analysis, and activity recommendations.
Project coordinator: Zbigniew Paszkiewicz, M.Sc.

The method is a contribution to both process mining area and computer supported collaborative work research area and will provide more efficient and effective operation of Virtual Organizations that are critical for global, knowledge-based economy. The proposed project concerns two main research areas: (1) computer support for Virtual Organization operation, and (2) process mining.

To face the challenge of fierce competition and demanding environment, small organizations base their operation on collaboration within Virtual Organizations (VOs). Virtual Organization (VO) is a set of at least two autonomous partners, where at least one of them is an organization, cooperating within a particular structure of social and legal relations in order to carry out a particular venture due to the demand from virtual organization clients and having a plan to carry out this venture. Effective computer support for such cooperation has been a subject of interest of many scientists. In particular, partner and service selection has been identified as a critical aspect of VO operation. In general, the methods proposed in this area take a form of predefined and structured processes. There is a need to propose an adaptive method of partner and service selection taking into account VO specificity and its particular context in the selection process. 

The method proposed in this project is a solution of this problem. The method permits adaptable and flexible partner selection by recommendation of actions. The method contributes to more efficient and effective selection of partners and services, and in this way – to more efficient operation of Virtual Organization and its comprising partners contributing to the development of economy.

Process mining is a research area having its roots in the following research areas: data mining, machine learning, business process modeling and analysis. The recent development of process mining was catalyzed by wide range of algorithms and methods elaborated in context of process discovery, conformance check and process extension, practical implementation of these methods and algorithms and development of standards. Recently a number of works has been initiated in the area of process prediction and recommendation. These areas having a potentially significant impact on efficiency of process executions have been identified as especially useful and promising. Mature methods concerning these areas are still to be proposed. The method proposed in this project contributes to this new and unexplored area of process mining by introducing the concept of activity patterns identification and recommendation, and context analysis.

The concept of the method that is a subject of this project is the next step beyond initial studies that have been conducted within other research project concerning Virtual Organizations and collaborative work support.

In case of successful implementation of the proposed method, it can be provided as a service by Virtual Organization Breeding Environment (VOBE, sometimes shortened to VBE in the literature). VOBE is defined as an association of organizations and their related supporting institutions, adhering to a base long term cooperation agreement, and adoption of common operating principles and infrastructures, with the main goal of increasing their preparedness towards collaboration in potential VOs. As a VOBE aims to increase preparedness of its members towards collaboration in potential virtual organizations, the proposed method contributes to this goal. 

The implementation of the proposed RMV method is performed as a part of the ProM Framework and will be integrated with existing ErGo System.

More about the research in the area of process mining at the Department of Information Technology on the http://www.kti.ue.poznan.pl/en/process.mining site.