Collaboration and Process Modelling in Engineering and Business - Methodology

Methodology

Management of ontologies and related models is essential for management and understanding of complex systems. This also improves the chances of project success. Breaking a complex systems engineering problem or project into modular sub-systems ensures that if failure looks likely it is possible to go back a step, rather than restart or abandon the project. This also informs, teaches and enables systems thinking and design. This can be achieved by transforming engineering knowledge into code in collaborative computer systems, and visualising the structure of the resultant system to all domain experts. This enables translation (both in human communication and computing model and program transformation) from the business and/or engineering focus of the domain expert end users, to the software developers and help desk and vice versa.

This translation can be enabled by encouraging end user visual programming so that the expertise of domain experts in modelling problems can be applied. Semantic Web modelling tools are needed to illustrate benefits this technology can provide to industry, as outlined at the Jena User Conference (Aragones et al., 2006).

Enabling, providing and supporting staff to use Semantic Web and Web 2.0 technologies can lower costs, improve collaboration, and improve staff satisfaction and retention. This enables a much faster iterative process of model development, visualisation and rework, leading to more effective process modelling. Such techniques can be used for managing and visualising workflows, which could involve complicated scheduling, sequencing, and iteration. Anderson (2007) in a Joint Information Systems Committee (JISC) report examines these advantages. Andersons’ JISC report explains how Semantic Web and Web 2.0 are related, as Berners-Lee's intention in the early development of Semantic Web technologies was for pages to be interactive. Andersons’ report talks of Web 2.0 trends towards the “End of the software release cycle, Lightweight programming models, Software above the level of a single device, and Rich user experiences.” The report explains how the technologies used are enabling user-centred web applications, and the use of the web as a development platform. Anderson advocates that "As a Web 2.0 concept, this idea of opening up goes beyond the open source software idea of opening up code to developers, to opening up content production to all users and exposing data for re-use and combination”. Anderson also establishes the need for communities to build ontologies. Software applications are needed that allow users with little software knowledge to edit and update ontologies themselves. Anderson talks of 'harnessing collective intelligence' by means of interactive collaborative software, he calls this 'distributed human intelligence'. So this kind of technology and these techniques are certainly appropriate to collaborative modelling, and sharing of knowledge amongst domain experts.

Visualisation of every stage in collaborative model creation and every part of the model enables determination of whether the model and view(s) of it are useful and correct, and so aids this sharing of human knowledge to solve problems.

A problem can be represented and shared as visual diagrams, and then translated from this abstract view to models and code. Gross (2007) argues the need for end user programming by designers using diagrams and scratchpads. Tufte (1990) explains how diagrams can be more effective than words for representing geometry. This links with the theme through this research of translating from an abstract to a concrete representation; Green et al. (2007) explain this distinction between abstract and concrete models. This distinction is more gradual than the distinction between classes and objects for object oriented programming. To achieve this it is necessary to enable translation from ontologies and Semantic Web information to diagrammatic representations and vice versa.

Green et al. (2007) discuss terminology problems, and how ontologies and models and collaboration can be used to help solve this problem. For complex modelling it is necessary to provide a generic model/system creator that would have the capability of translating abstract ideas into models. This provides a translation stage that closes the gap between ideas and models. A classification scheme or ontology is necessary in order to make communication as precise as achievable. Ontology/ies can be used to define terms and higher level relationships and these extended for application to particular modes. Such ontology/ies can also be used to help non-specialists understand the terminology of a particular domain, and so ease collaboration amongst people with very different skills and backgrounds. For this project a web linked modelling environment is constructed that automatically translates an ontology representation to models, performs calculations/inferencing, and assists decision support.

Sternemann and Zelm (1999) explain that it has become necessary to research collaborative modelling and visualisation tools, because of the business trend towards global markets and decentralised organisation structures; Green et al. (2007) also explain this. This research aims to benefit two types of modeller. Firstly, it is intended to create a system for modellers to develop models from shared information (the process architecture and models repository); such 'model builders' will be able to create process models for ‘model users’ who may customise models but will not need to create them. In practice, an individual may be in either, or both, categories depending on their interests and skills. In effect this involves production of systems to create systems and/or models.

McGuinness (2003) explains how ontologies support this functionality and calls this ‘generalization/specialization’ of information. Uschold and Gruninger (2004) describe ‘ontology-based search’ as “Ontology used for concept-based structuring of information in a repository”; and describe the benefit of this as “better information access” This aids the research objectives of ease of use and sharing of information. The use of open standards for representing information makes it possible to enable searches that understand the semantics of the information and so can track all of the relationships between items. To test this methodology web based example models were created.

As a shared ontology is central to the proposed system, a translator or an 'elaborator/de-abstractor' can be developed that translates and converts, a diagrammatic representation of the problem automatically into code. Translations could be performed in any programming or meta-programming language, or open standard representation languages, the visualisation of the model created being suitable for display on the web. Interactivity is the key to human involvement in the Semantic Web, in order to construct a process modelling system for people in industry who are not professional programmers.

Figure 1 shows the theory behind Semantic Web modelling. These modules/sub-systems have been prototyped in order to verify that the methodology is feasible.

1. Connections are established between the ontology system and databases, spreadsheets, or other systems that hold relevant information for that modelling problem.
2. The ontology is created using RDF/OWL (Miller and Baramidze, 2005) and an interface built to edit it.
3. Libraries are created in a partnership between ourselves and domain experts.
4. Taxonomies/sub-ontologies are populated by model builders who use them for modelling problems. These are based on the libraries created in step 3.
5. Taxonomies are colour coded for ease of understanding. A link was created between the ontology tool and the illustrated process modelling system.
6. There are two sorts of constraints that can enable users to build and adapt models. Constraints on the way the ontology, and models are built, and user interface constraints to reduce the scope for error, (this is not yet developed).
7. Colour coding makes inferencing/sequencing clear because all taxonomies can be used in any equation, inferencing outputs of a multicoloured result tree that represents the entire sequencing system. User choices affect how items are related for the equation; choices could be made manually or via a search.
8. Each node represents uncertainty. Uncertainty depiction within inferencing results was prototyped (Bru et al., 2004), and used in the DATUM project.
9. The result tree is represented on the web and in other programs allowing further searching, processing and evaluation of results. Visualisation and use of searchable languages such as XML, and SVG (Scalable Vector Graphics) can assist this.
10. and 11. Modellers interact with the ontology, the model, and results. There is a two way feedback mechanism where experts can make changes in any sub-system, and this filters into changed results. This supports a cycle of results and rework.




Figure 1 - Semantic Web based Process Modelling system

If users are involved in construction and/or use of such sub-systems and collaboration is organised so knowledge is transferred between those involved, the rational for process models and decisions can then be understood.

Translation involves the elaborator that follows a structured ontology to establish how related concepts (represented visually) can be represented in equivalent code and vice versa. This elaboration process converts an abstract representation of a problem to a concrete model created with the aid of communication between user and modelling tool. The translation program calculates results, and these are translated back to users. Results are visualised using stylesheets and interactive software, and translated between different representations. Information must be visualised in the way most appropriate to particular modellers according to their situation and the problem type examined. Highly interactive Semantic Web applications that provide a user interface can enable a user centred process modelling system.
The methodology was tested by creating a practical system. Figure 2 shows the 3 step translation process from domain level visualisation to code, via the web and/or programming languages, and resultant visualisation. To allow information sharing, reuse, interoperability and collaboration, an ontology centric approach (Bloodsworth and Greenwood, 2005) is used, which acts as the infrastructure for creation of models and model templates.


Figure 2 - Translation Process Implementation

To build on this test system a visual user driven process for model development is evaluated, which merges approaches of object oriented design, the Semantic Web, relational databases, event driven programming, model driven programming, and visual programming resulting in application of 'The Model Driven Semantic Web' (Frankel et al., 2004). Further, we examine ways of structuring information and enabling processing and searching of the information to provide a modelling capability, thus providing a solution for the key issue of increasing user involvement in software. To achieve this, visual templates are constructed for the modeller to enable non-programmers to develop modelling software. Increasing user involvement in software development is recommended by Olsson (2004), the involvement being essential to avoid communication problems and misunderstandings that result from domain expert users having too little influence on tool creation.


References

Anderson, P. Technology and Standards Watch (2007). JISC (Joint Information Systems Committee) What is Web 2.0? Ideas, technologies and implications for education [online]. Available from: http://www.jisc.ac.uk/media/documents/techwatch/tsw0701b.pdf [Accessed 24th July 2009].

Aragones A, Bruno J, Crapo A, Garbiras M, 2006, An Ontology-Based Architecture for Adaptive Work-Centered User Interface Technology, Jena User Conference, Bristol, UK.

Bloodsworth P, Greenwood S, 2005, COSMOA: An Ontology-Centric Multi-Agent System For Coordinating Medical Responses To Large-Scale Disasters, AI Communications 18.

Bru, C., Scanlan, J., Hale, P., 2004. Visualization of Cost Information, International Journal of Agile Manufacturing, 7(1), pp 53-59.

Frankel D, Hayes P, Kendall E, McGuinness D, 2004, The Model Driven Semantic Web, MDSW2004.

Green, S., Beeson, I., Kamm, R., 2007. Process architectures and process models: opportunities for reuse. In: 8th Workshop on Business Process Modeling, Development, and Support BPMDS07 and CAiSE'07 11-15 June 2007, Trondheim, Norway.

Gross, M. D., 2007. Designers Need End-User Software Engineering. In: End-User Software Engineering Dagstuhl Seminar February 2007.

McGuinness, D. L., 2003. Ontologies Come of Age. In: Dieter Fensel, Jim Hendler, Henry Lieberman, and Wolfgang Wahlster, ed. Spinning the Semantic Web: Bringing the World Wide Web to Its Full Potential. MIT Press, 2003.

Miller J A, Baramidze G, 2005, Simulation and the Semantic Web, 2005 Winter Simulation Conference.

Olsson E, 2004, What active users and designers contribute in the design process, Interacting with Computers Vol 16.

Sternemann, K. H., Zelm, M., 1999. Context sensitive provision and visualisation of enterprise information with a hypermedia based system, Computers in Industry Vol 40 (2) pp 173-184.

Tufte, E., R., 1990. Envisioning Information. Graphics Press.

Uschold, M., Gruninger, M., 2004. Ontologies and Semantics for Seamless Connectivity. In: Association for Computer Machinery - Special Interest Group on Management of Data - SIGMOD Record December, 33(4).