Ontology learning; Qualitative Analysis; Environmental management; case study research; grounded theory; qualitative research; ontologies; theory building; coding; taxonomy; global report initiative
Han Dong, Stoffel Kilian (2013), Ontology based Model and Procedure Creation for Topic Analysis in Chinese Language, in Juanzi Li (ed.), Springer, New York, 67-73.
Zingg Aline, Reiner Gerald (2013), The value of voluntary sustainability reporting - an event study in the financial services industry, in Proceedings of 20th EurOMA Conference
, Dublin, Irlande.
Han Dong, Stoffel Kilian (2012), An Interactive Working Tool for Qualitative Text Analysis, in 12th French-speaking Conference on Knowledge Extraction and Management
Han Dong, Stoffel Kilian (2012), Ontology based Qualitative Methodology for Chinese Language Analysis, in Advanced Information Networking and Applications Workshops (WAINA)
, Fukuoka, Japan.
Gomes Dos Santos Gil, Reiner Gerald (2011), A method for assessing environmental management in financial service firms, in 4th International EurOMA Service Operations Management Forum
Gomes Dos Santos Gil, Han Dong, Reiner Gerald, Stoffel Kilian (2011), Method for Screening Sustainability Reports in the Financial Services Industry, in 14th Toulon - Verona Conference
, Alicante14th Toulon - Verona Conference, Spain.
Han Dong, Kilian Stoffel (2011), Ontology based Qualitative Case Studies for Sustainability Research, in Proceedings of Workshop "A.I. for Intelligent Planet", IJCAI 2011
, BarcelonaIJCAI, SPAIN.
Qualitative case study research have gained more and more importance for theory building for the management domain. It can be used to bridge rich qualitative evidence to mainstream deductive research, but the analyzing data phase, which is the heart of building theory from cases, it is the most difficult as well as least standardizedpart of this process. Consequently, this approach is difficult to be applied to complicated but very important issues, as the comprehension of the environmental orientation at firms. It is absolutely necessary to overcome the deficiencies of the case study research, i.e. the lack of objectivity. At the same time ontology based reasoning has become a very active research field in the domain of computer science. Ontologies are the specification of a conceptualizations, using share knowledge in human and machine readable form. However the creation of ontologies, especially for a domain expert not necessarilytrained in formal methods, is still a big challenge. The goal of this interdisciplinary project is, on the one hand, to make a contribution in the field of case based studies by adding ontologies into the overall process, and on the other hand to integrate the approach of building theories from case studies to the field of ontologies in computer science. The resulting approach is validated using a real case study research, projected to develop overall understanding of environmentalmanagement orientation and their impact on financial service firms performance. In order to achieve this goal we will first develop a formal model for case study research based on grounded theory. This model will be presented in form of an ontology consisting of three layers. The first describes all formal aspects of grounded theory based case studies, namely the building of codes, concepts, categoriesand theories. At the same time this ontology will define the reasoning mechanisms which adds the requested rigor. The second layer will add the specification of given case study project as well as the expert knowledge needed for that project. The third layer will be produced by the researcher running the case study. The ontology approach will allow to verify that his work is conform with the project, expert, and grounded theory ontology. Based on this model, a prototype implementing the ontology layers and supporting supplementary functionalities will be created and used as working tool for the case study in environmental management. An important practical issue for this prototype is the creation of the coding ontology (the third layer). This ontology must be learned from raw data, expressed as unstructured documents and annotated by the user during the coding process. The input data type oriented our choice towards a variant of LDA model, called label-topic model, for which the labels(codes) and documents are represented as random mixtures over latent topics, where each topic is characterized by a distribution over words. Because each label (concept in the coding ontology) is expressed as a distribution, statistical divergence measures can be used to generate a hierarchical structure (taxonomy). Moreover, this probabilistic model allows the evaluation of the qualityof the coding process by using the perplexity measure. And because the trained model can predict the topic distribution for a new, not yet annotated document, it can be used to implement a recommendation system providing the ”best” code for this document (codes with the highest posterior probabilities). The main purpose of the case study project is to develop a better understanding of the environmental processes at selected financial service firms from a service operations management perspective. This is done by dealing with the following major research questions: 1. What are the reasons for environmental management orientation? 2. How is the implementation conducted? 3. Which mode of environmental management orientation is chosen? 4. What is the performance impact of different environmental management orientations?.Our case study research is based on multiple years of standardized environmental reports (GRI, global report initiative) to develop theories in this context and to validate them. First, we will formulate a clear research framework to describe the link between cause and effect. These works will provide later on the input for our concepts. In general, we are applying an ontology consisting of three layers to enable/facilitate theory development.