Gregory Richards

Toward a Goal-Oriented, Business Intelligence Decision-Making Framework

Decision making is a crucial yet challenging task in enterprise management. In many organizations, decisions are still made based on experience and intuition rather than on facts and rigorous approaches, often because of lack of data, unknown relationships between data and goals, conflicting goals, and poorly understood risks. This paper presents a goal-oriented, iterative conceptual framework for decision making that allows enterprises to begin development of their decision model with limited data, discover required data to build their model, capture stakeholders goals, and model risks and their impact. Such models enable the aggregation of Key Performance Indicators and their integration to goal models that display good cognitive fit. Managers can monitor the impact of decisions on organization goals and improve decision models. The approach is illustrated through a retail business real-life example.

Towards outcome-based regulatory compliance in aviation security

Transport Canada is reviewing its Aviation Security regulations in a multi-year modernization process. As part of this review, consideration is given to transitioning regulations where appropriate from a prescriptive style to an outcome-based style. This raises new technical and cultural challenges related to how to measure compliance. This paper reports on a novel approach used to model regulations with the Goal-oriented Requirement Language, augmented with qualitative indicators. These models are used to guide the generation of questions for inspection activities, enable a flexible conversion of real-world data into goal satisfaction levels, and facilitate compliance analysis. A new propagation mechanism enables the evaluation of the compliance level of an organization. This outcome-based approach is expected to help get a more precise understanding of who complies with what, while highlighting opportunities for improving existing regulatory elements.

Creating Quantitative Goal Models: Governmental Experience

Precision in goal models can be enhanced using quantitative rather than qualitative scales. Selecting appropriate values is however often difficult, especially when groups of stakeholders are involved. This paper identifies and compares generic and domain-specific group decision approaches for selecting quantitative values in goal models. It then reports on the use of two approaches targeting quantitative contributions, actor importance, and indicator definitions in the Goal-oriented Requirement Language. The approaches have been deployed in two independent branches of the Canadian government.

A Framework for a Business Intelligence-Enabled Adaptive Enterprise Architecture

The environments in which businesses currently operate are dynamic and constantly changing, with influence from external and internal factors. When businesses evolve, leading to changes in business objectives, it is hard to determine and visualize what direct Information System responses are needed to respond to these changes. This paper introduces an enterprise architecture framework which allows for anticipating and supporting proactively, adaptation in enterprise architectures as and when the business evolves. This adaptive framework exploits and models relationships between business objectives of important stakeholders, decisions related to these objectives, and Information Systems that support these decisions. This framework exploits goal modeling in a Business Intelligence context. The tool-supported framework was assessed against different levels and types of changes in a real enterprise architecture of a Canadian government department, with encouraging results.

Regulation-Based Dimensional Modeling for Regulatory Intelligence

Regulations are a source of evolving requirements for products and organizations. As regulatory institutions shift towards outcome-based regulations, they increasingly adopt legislation performance modeling, at the basis of regulatory intelligence. In this context, performance modeling refers to the measuring of important business aspects in a coordinated manner and the use of these measurements for improved decision making. Considering that in many cases regulatory texts already exist, it is necessary to build a performance model based on existing regulations that may be still prescriptive rather than outcome-based. The process of turning the underlying textual legislation into a formal performance model that can be assessed by Business Intelligence (BI) tools is complicated due to organizational, cultural, and technological reasons. In this paper, we present a methodology from a technical perspective that enables regulatory institutions to reason about regulations and compliance with regulations as new dimensions. We demonstrate the methodology using traffic law as an example regulation, jUCMNav for performance modeling, and IBM Cognos for BI reporting.

Information Technology Artifacts in the Regulatory Compliance of Business Processes: A Meta-Analysis

To address recent calls for better regulatory compliance of business processes (electronic or not), this study synthesizes research findings of nine peer-reviewed systematic literature reviews and state-of-the-art studies on business process compliance to examine Information Technology artifacts used in the compliance of business processes to laws and regulations. Results from the 342 unique studies covered by these reviews demonstrate that eight types of Information Technology artifacts are used in the design, execution and after-execution lifecycle phases of regulatory compliance to perform compliance modeling, checking, analysis and enactment tasks. The results also demonstrate with statistical evidence that compliance enactment (which involves putting mechanisms in place to address changes in the compliance representations, business processes or regulatory requirements) has received the least attention. This is a worrisome situation given the dynamic environment in which businesses operate. We argue that the findings in this work are relevant toward understanding the current state of the domain, providing insights on what phases and tasks need improvement.

The Road to Decision-Centric Business Intelligence

Business Intelligence (BI) is increasingly becoming a decision-making tool of great importance to organizations. However, BI–mediated information delivery is driven more by data models than by “management” decision models. The purpose of this exploratory study is to explore the possibility of modeling management decisions as a first step towards the development of decision-centric BI tools. The methodology is interpretive qualitative research comprising interviews with six senior managers and executives. Findings suggest that decisions featuring less uncertainty with measurable outcomes are more amenable to the use of decision models as a precursor to a decision-centric BI-mediated information delivery.

Benefits and Barriers to Corporate Performance Management Systems

Corporate performance management (CPM) systems using business intelligence technologies can help enterprises monitor and manage business performance. In this research, we explored and presented empirical evidence on the key benefits of, and barriers to, the use of CPM systems through a survey of 283 organisations across North America and China. We identified three key benefits (strategy execution, process efficiency, and fact-based decision-making) and ten inhibiting barriers under respective project and organisational dimensions. Moreover, we found that in regard to the use of CPM systems, Chinese organisations perceived higher benefits, as well as higher barriers, than did their counterparts in North America. The socio-cultural differences between the two regions explain these issues. The research findings are useful for multinational organisations that are planning, or are in the process of implementing or reviewing their CPM systems, as well as for consulting companies that are assisting with such systems implementation in different regions.

Drafting and modeling of regulations: Is it being done backwards?

The performance modeling of regulations is a relatively recent innovation. However, as regulators in many domains increasingly look to move from prescriptive regulations towards more outcome-based regulations, the use of performance modeling will become more common place. The major difference of outcome-based regulations over prescriptive regulations is that the main interest lies in specifying clear objectives of the regulations and measuring whether regulated parties achieve these objectives, while leaving much freedom to the regulated party on how to meet these objectives. Recently, we have found that the use of performance modeling provides benefits such as revealing inconsistencies and lack of clarity in existing regulatory language. In this paper, we report on these experiences, summarize guidelines for the modeling of regulations, and examine whether the current drafting processes for regulations are optimized to take advantage of these additional benefits. We explore the advantages and disadvantages of various ways of augmenting the current approach with goal-oriented modeling of regulations. Based on our experience with Aviation Security regulations, we believe it is time for modeling to play a new role in helping to guide the drafting of regulations.

A distributed service oriented infrastructure for business process management in Virtual Organizations

Virtual Organizations (VO)-a network of independent organizations collaborating to address specific business opportunities-have become popular in todays technology driven business environment business climate. Although many solutions have been proposed over past two decades to support inter-organizational interactions, VOs still face several challenges due to their dynamic and temporal nature that cannot be addressed by the traditional solutions such as ebXML and EDI. In this paper, we propose a distributed architecture for a flexible infrastructure that supports collaborative business process execution, monitoring and management. The infrastructure supports the common topologies of inter-organizational collaboration (e.g. peer-to-peer, star, linear), and responds well to changes due to its loosely coupled components. The proposed infrastructure is based on federating multiple SOA infrastructures with the Enterprise Service Bus (ESB) as its core. It is implemented using IBM WebSphere products. We demonstrate that the proposed SOA-based virtual organization facilitates flexibility, rapid process integration and dynamic evolution.