Back to top

Master Thesis Abm Junaed

Last modified May 24, 2020
  2015

 

Abstract

Complex products such as commercial aircraft and the associated development activities require complex engineering processes which are supported by IT tools. The IT landscape is characterized by a set of very heterogeneous engineering tools providing data in different formats, e.g., relational databases, XML, CSV, XLS, and alike. However, there is no single API to access the data. A key approach at Airbus Group Innovations (AGI1) to tackle this problem is the application of Linked Data and Semantic Web technology. Web Ontology Language (OWL) is one of the core components of the Semantic Web stack, which is used for knowledge representation (KR). But, OWL also comes with a learning curve for the domain experts who have little or no knowledge in ontology engineering. To minimize the learning curve to access such data, the primary purpose of this thesis is to provide a solution consisting of a web-based natural language interface (NLI) to manage complex product ontologies. The purpose of the NLI is to guide the domain experts to create and update OWL ontologies, as well as to search the ontology to find out the inconsistencies or missing concepts. An additional contribution of this thesis is to provide a mechanism to reuse existing ontologies through the NLI.

We apply a design science methodology to tackle the challenges. A state of the art study was conducted to detect a useful basis for our solution. As part of the solution, a web-based prototype is presented, which uses an NLI. This NLI is based on a CNL and a Semantic Wiki. Moreover, our prototype can reuse existing ontologies by importing them into the system and by creating domain specific lexicons automatically.

The prototype was evaluated in several ways. Qualitative interviews with Airbus stakeholders yielded valuable feedback. Moreover, several real ontologies of an aircraft were used to carry out functional tests and these tests showed promising results. The prototype is also highly portable to different OWL ontologies, since it does not require any customization.

 

Research Questions
The research questions which will be tackled in this study are grouped into two different
categories:
1. Major research questions: Primary focus of this thesis will be given towards the
following research questions:
1.1. How to create an OWL ontology using a web-based NLI?
1.2. How to search in OWL ontology using a web-based NLI?
1.3. How to incorporate existing ontologies into the proposed NLI?
1.4. How to create domain specific lexicon automatically from existing ontologies?
2. Derived research questions: Following research questions are derived from the
challenges associated with NLIs:
2.1. How to guide the user to formulate queries, to add and edit data into knowledge
base by providing minimum training while keeping the supported language
intuitive?
2.2. How to resolve the ambiguity of natural language?
2.3. How to keep the NLI portable?
2.4. How to hide the underlying complexities of the structured knowledge from the
end user?

Files and Subpages

Name Type Size Last Modification Last Editor
Junaed_final_presentation.pdf 2,07 MB 11.07.2016
Junaed_final_presentation.pptx 2,46 MB 11.07.2016
Junaed_kickOff_final_14_12_2015.pptx 3,44 MB 29.06.2016
Junaed_Thesis_Final.pdf 2,90 MB 11.07.2016