Managing extended organizations and data governance : a foucauldian perspective Eric Buffenoir, CNRS, Montpellier, France Isabelle Bourdon, MRM, Université Montpellier 2, France Digital Enterprise Design & Management (DED&M) 2013 February 12-13, 2013- Paris, France
Issues Bad data = 700 billions $/ year (Ovum 2010) Data quality for control management and business intelligence. 2 paradigms: Master Data Management and Community Management. Specific issues of data governance schemes for extended organizations?
Plan Data quality management / Data governance Data management issues within Extended organizations The existing data Governance paradigm s : principles and limits Guidelines of a data governance scheme devoted to extended organizations Questionning data governance issues from a Foucaldian standpoint Claim for a new data governance paradigm
Data management issues Data governance = rights and duties concerning definition, distribution, synchronization and exchange of reference values for Master Data Data governance scheme necessitates : semantic alignment of business glossaries & standardization of business processes, identification of control authorities/ranked sources distribution of prioritized roles and responsibilities.
Extended organizations
Data management issues in extended organizations Data governance schemes must promote : Decentralization of control processes over Master Data Integration of inter- or trans-organizational communities and partner organizations in data governance Development of transverse mechanisms and tools, Implementation of dictionaries between organization s Master Data and norms associated to external sources.
Master Data Management
Community-based data management
Data management in extended organization
Normalization after Michel Foucault
M.Foucault and Information Technology
Claim for a new data governance paradigm
We can summarize the guidelines of a new data governance paradigm adapted to extended organizations : This paradigm is user centric, the user interfaces offers simultaneously access to services and rights on a selected collection of data associated to them. This scheme covers the set of Master Data and Hierarchical data involved in a progressively growing set of numeric services offered to users. The data are presented in their current state of reliability, the system maintains its reference databases through real-time processes. Complete transparency and traceability are ensured on the set of required interventions made from individual actions or from external sources. A control channel is associated to any community concerned by a specific common usage relied on any given datum. Each control channel is formalized by the allocation of structuring roles and prioritized rights concerning this datum to any individuals within. The set of control channels aims to encompass the whole set of relations inherited from networks existing within the organization. Modifications made by individuals on shared hierarchical data contribute to change their field of action and the area of visibility they are placed in, through competitive processes. Control channels concerned by the same collection of data are ranking/censoring/granting each other according to the rights they have to act on hierarchical data. A unique control channel associated to hierarchical channel inherits the arbitration/censorship power on a given datum The system promotes a "functional decentralization" of a significant part of the control processes through the development of a distributed IS architecture based on numerous instances of this IS brick. The next step is to develop a platform that implement those propositions.
Thanks Eric Buffenoir, CNRS, Montpellier, France Isabelle Bourdon, MRM, Université Montpellier 2, France Contact : isabelle.bourdon@univ-montp2.fr