Proceedings of the 4th Annual Sessions 31st October 2007 

Using Human-Assisted Machine Translation to overcome language barrier in Sri Lanka

Budditha Hettige, Asoka. S. Karunananda

Abstract-Automated machine translation has faced with many issues regarding handling of semantics. In general, this issue can be addressed by computer-assisted machine translation at the pre-editing and post-editing stages. Our research has gone further and introduced an intermediate editing stage just before morphological analyzer of the target language. This approach detects semantic issues before hand and allows addressing those by human intervention. As a result the final translation will be more realistic and cut down the need for human intervention at the post-editing stage. The above approach has been used to develop English to Sinhala machine translation system. The system has been developed using Prolog and Java to run on a standard PC.



On Computing Ontology for Mental Factors

Subha D. Fernando & Asoka S. Karunanada

Abstract-Ontology has introduced a new paradigm for software development. Nowadays, Ontologies has become a key theme of areas like Semantic Web, Web Services and Multi Agent Systems. However, a few research projects are carried out to develop a comprehensive Ontology for modeling of mental factors. This has been a barrier for modeling of computer systems concerning human emotions and sensations. We have exploited Buddhist theory of mental factors and developed a Mental Ontology environment to model systems involving mental factors. Our implementation introduces a base class mental factor with various feature and behaviors as per Buddhist theory of mind. This base class can be extended to create an arbitrary mental factor. The paper also explains how the developed mental Ontology environment can be used by an Ontology developer in Multi Agent System.



Using an Intelligent Agent with PBL approach to Online tutoring and mentoring

Udayanthi Weerasoooriya & Ajith P. Madurapperuma

Abstract-Online Tutoring and Mentoring is an essential activity for effective e-Learning systems. However, at present Online tutoring/ mentoring is conducted merely on the basis of personal experiences of tutors and mentors, without a theoretically-based approach. This paper presents an approach that exploits the theory of Problem-Based Learning (PBL) for online tutoring and mentoring. The proposed approach drives the online tutoring/mentoring process through a four dimensional framework concerning; identification of facts, generating ideas, identifying learning issues and preparing an action plan. In our approach, online tutor/mentor is simulated as Agent software that runs on a Learning Management System of an e-Learning environment.



NLP – Based Expert System for Database Design and Development

U. Leelarathna, G. Ranasinghe, N. Wimalasena, D. Weerasinghe & Asoka S., Karunananda

Abstract-Database designing and development involves a sequence of tasks including extracting the requirements, identifying the entities, their attributes, relationships between the entities, constraints, drawing a conceptual schema, mapping the ER diagrams to the database schema, and eventually developing the database. As such database design and development has become a tedious task for novice person. In addressing the above issue, we propose a Natural Language Processing enabled Expert Systems, which accepts textual domain descriptions and generate a relational database schema followed by a database. The entire NLP enabled system can be customized to link up as a front-end for any database management system. The system has been developed using Prolog, Flex and C#.net.



Computational modeling in conceptual models: Widening scope of artificial life

D.S. Kalana Mendis, Asoka S. Karunananda & U. Samarathunga

Abstract-Artificial Life or Alife is concerned with computational models invented by considering explicit knowledge in biological systems. However conceptual models have not been addressed for Alife due to involvement of informal practising methodologies. This leads to concern with tacit knowledge in conceptual models. This paper presents a research, which is incorporated with a computational modelling in diagnosis of human constitutions in Ayurvedic medicine considered as conceptual model. Further Tacit knowledge is the key issue of knowledge modelling aspect because all knowledge is rooted in tacit knowledge. An Intelligent Hybrid system involved with artificial intelligent techniques, namely fuzzy logic and expert system technology has been used to implement the computational model. The result of the modelling of Ayurvedic domain using fuzzy logic has been compatible with the experiences of the Ayurvedic experts. It has shown 77% accuracy in using the tacit knowledge for reasoning in the relevant domain.



Ontology driven approach to Disaster Management

A. Thushari Silva & Ajith P. Madurapperuma

Abstract-Disasters have become a common threat to humans. Although disasters cannot be evaded, human can manage the disastrous situations to minimize the damage. A successful disaster management system requires information from heterogeneous data sources that belongs to various services. As such, data integration is a crucial issue for disaster management systems. Recent research has recognized ontology as a potential approach to develop solutions for systems involving diversified sources of data. This paper presents an ontology driven approach to data integration in disaster management systems. The proposed ontology enables editing, sharing, converting, searching and querying on data available in various formats in databases, web servers, text files and image files. Protégé ontology development environment has been used to implement the proposed disaster management ontology.


Intelligent Elevator Group Control System

T.K.W. Senevirathna, H. Ekanayake & T.N.K. De. Zoysa

Abstract-High rising building is a common sight in most of the cities today. Fast and efficient elevator transportation is a key feature when creating these kinds of buildings. Lot of research has been carried out to build an intelligent system that satisfies the need of elevator control as it is a must to have intelligent elevators in the future. This paper proposes a destination control system for elevator group controlling which fully utilizes destination information. Fuzzy logic concepts are used to enable the elevator control system to make decisions. The design criteria include of optimizing movement of elevators with regard to several factors such as waiting time, riding time, energy, load, etc. Software simulation is done in order to capture the performance of the proposed system with compared to conventional approaches.