Titulo Estágio
ESOTERIC: Exploring domain specific industry verticals for USDL
Área Tecnológica
Sistemas de Informação
Local do Estágio
GSI/DEI-FCTUC
Enquadramento
USDL (the Unified Service Description Language) is gaining momentum as an approach to describe business services and cloud services. This movement is now clear and it was confirmed in a recent report requested by the German Federal Ministry of Economics and Technology (BMWi) on “Standardisation in Cloud Computing”. The report, released in February 2012, states that “the potential contained in a large number of established standards (e.g. USDL, OAuth, SCAP, or WS-*) … are becoming important for cloud computing.” Furthermore, the use of USDL in many European projects (e.g. FI-WARE, SOA4ALL, RESERVOIR, SHAPE, etc.), its standardization process with W3C (http://www.w3.org/2005/Incubator/usdl/), the creating of a community of practitioners, and the encoding of USDL as an RDF ontology (http://linked-usdl.org) indicate that it will most likely become mainstream in less than 18 months.
In its present form, USDL is a general purpose service description language (represented as a metamodel or ontology) which has to serve a wide range of industries, enterprises, and functional areas in a standardized manner. However, general purpose languages are often not expressive enough to capture the specificities of domain specific industry verticals. The generalized application of a general purpose service description language comes at the expense of a greater complexity when customization is needed in order to model domain specific services from, for example, the telecommunication, healthcare, and energy sectors. As such, USDL faces several challenges and difficulties ahead, such as complex configuration processes, industry mismatch and low adaptation to particular industries. In order to alleviate these potential drawbacks, one approach we want to pursue and explore is to look into how USDL can be adapted or extended to capture the singularities of domain specific industry services. This is the goal of this project called ESOTERIC (Definition of esoteric: Intended for or likely to be understood by only a small number of people with a specialized knowledge or interest.)
Objetivo
Most industries have particular requirements and a general purpose service description language, such as USDL, will most likely not serve as a “one size fits all” approach. USDL can be seen as a canonical model which will not be able to “please all of the people all of the time”. While the language is suitable to offer a high level description of a business or cloud service, it will have difficulties to fit with firms’ specific activities, vocabularies, business models, business processes, among others. Thus, these limitations will force firms to undertake the customization and configuration of the language with local, closed, proprietary extensions and adaptations to meet business needs. This customization will have considerable development and maintenance costs.
For example, in the telecommunication industry, how can the notion of Router, Device Interface, Value Network Role Entities, Supplier/Partner Plan, Project, Trouble Ticket, and Usage be modeled with USDL? It is clear that customization work is required by firms to meet specific market needs. Nonetheless, the availability of a family of domain specific industry vertical USDL modules would be obviously a better choice for a telecommunication company rather than using the general purpose USDL language. This scenario is the main trigger for this research proposal and seeks to study how the USDL specification can be customization to also model domain specific industry vertical services.
The idea to create industry-specific USDL modules is not new. For example, in the field of ERP (Enterprise Resource Planning) systems, SAP, one of the leading worldwide vendors, defined a similar strategy which was based on delivering customized enterprise systems across industries such as manufacturing, insurance, banking, finance, healthcare, telecom, energy and utilities. Although USDL can become a building block for several service industries, research on extending and/or adapting USDL is important for service-based economies and has not yet received the needed consideration. To advance research in this field, we propose to explore how USDL can be tailored for specific service industries and evaluate the dichotomy between model customization and complexity. The rational is that an excessive customization or over engineering can lead to a lower cost-benefit ratio, while a high level of generalization cam lead to an inadequate modeling power.
Plano de Trabalhos - Semestre 1
(a) Discussion, collaboration and research with KMI/The Open University (UK) and SAP Research (Germany) on the industry domain to select to study and for which a metamodel/ontology will be developed. Business use case definition which will drive the research project until completion (September de 2012)
(b) Specification of the requirements (technical, functional, environmental, industrial, etc.), tools, languages, models, and systems to use in the project (October a November de 2012). Establish a dependency structure which expresses the relationships of ESOTERIC with other efforts within the ISG research group. Identify synergies.
(c) Architecture, design and modeling (December 2012 and Janeiro 2013) – Present a detailed analysis and definition of the modeling requirements for the domain specific language for USDL to develop. Provide a design of the supporting/enabling system. Using rapid prototyping, illustrate the innovative concept of domain specific industry verticals for USDL.
(d) Writing and defense of mid-term report (December 2012 and Janeiro 2013).
(e) First prototype (February de 2013). Generate the first demo version of the system articulating the concept of domain specific industry verticals for USDL. Show it can be effectively used with the business use case definition (step a)). The prototype will help defining additional functional requirements and will also allow receiving feedback from the member of the group. Relate the vertical extension with USDL core.
Plano de Trabalhos - Semestre 2
(f) Construction (February and May 2013) – Programming and modeling the components, modules and ontologies defined. Second version of the business use cases to better illustrate the benefits of domain specific industry verticals for USDL.
(g) Testing (April and June 2013) – Definition and execution of the acceptance tests of the various modules and ontologies developed. Receive and address feedback from KMI/The Open University and SAP Research.
(h) Documentation (running during the project) – Integration and revision of the documentation produced in each phase of the project into the final documentation which describe the work carried out technically and from a business perspective. Write the final version of the business use case.
(i) Writing and defense of the final report (May and July 2013). Final master thesis writing, coding, integration, testing, modeling, cleaning, etc. If time allows, take the final master thesis and formatted it with a suitable format to be submit to a conference. (Help will be provided).
Condições
This work will be carried out at DEI/Universidade de Coimbra within the Information System Group (ISG). A suitable space will be made available to students. Meetings will be held regularly (every week) with several students working on the business and cloud service descriptions using USDL and Linked-USDL.
Observações
[Definition] Esoteric: Intended for or likely to be understood by only a small number of people with a specialized knowledge or interest.
Orientador
Jorge Cardoso
jcardoso@dei.uc.pt 📩