http://dlib.iit.ac.lk/xmlui/handle/123456789/135 Tue, 21 Apr 2026 06:27:45 GMT 2026-04-21T06:27:45Z http://dlib.iit.ac.lk/xmlui/handle/123456789/212 Trend Predictor Fashion trend prediction tool for businesses Sheriff, S Companies spend millions sometimes on products which don’t get sold. For companies, which solely manufacture apparel, it becomes a huge issue as they don’t get to sell these items and when stock gets returned they have to find space to store them as well. Which in turn makes companies buy warehouses for storage or additional surplus charges that they have to adhere. As items, don’t get sold or when they are returned companies suffer bigger and bigger loses. This document presents the reader with a high-level description of the project. An automated system for predicting the next trend for the next 6 months or for a year. It encapsulates the problem that the software is trying to address and the feasibility of the proposed project. The document also contains the anticipated schedule of deliverables and moreover, the resources that would be required to successfully proceed with the project. The project aim and objectives are also highlighted herein. The aim of this project is to develop “Trend Predictor”, a dedicated trend prediction system, which lets designers of manufacturing companies to get predicted data on trends and associate them with the current trends, thus allowing companies to produce apparel fitting the future market. Sun, 01 Jan 2017 00:00:00 GMT http://dlib.iit.ac.lk/xmlui/handle/123456789/212 2017-01-01T00:00:00Z http://dlib.iit.ac.lk/xmlui/handle/123456789/211 Driving behavior analyzer Sumanaratne, S. The driving behavior is a key feature which all the drivers are concerned. Through knowing his/her own driving styles, it will be benefitted to save the wasting travel time, fuel and the safeness of the driving. Driving behavior not only depends on the drivers’ usual driving styles, but also environmental factors. Based on the vehicle type and the driving experience also may be effected to change the driving behavior of a person. Specially, when considering the one person’s driving styles in different vehicle types, it will definitely show behavioral variations, because more driving experience in a particular vehicle type is needed to control the vehicle in any condition. When examine the driving behavior, according to the road types also persons’ driving style can be changed. Therefore, having a satisfying knowledge in own driving styles in different road, vehicle or environments condition will be helped improve their own driving skills. After considering the above mentioned features, as a solution for identifying different driving behaviors a software solution has been proposed. The system will detect the drivers’ behaviors in different road types and according to the road type best performance will be selected. Also the proposed system will allow users to view their own driving styles in line graphs. So the users will be able to view how their driving behavior has been changed along with the travelled time. Also the map of the driving route can be load through this system. This solution is implemented based on the Classification methods. Artificial Neural Network (ANN) is used as the main algorithms and the Multilayer Perceptron is used as the classifier for the proposed solution. Because of having limited time frame some feature were developed to a certain extent and rest of the features will be developed in the next development phase. Sun, 01 Jan 2017 00:00:00 GMT http://dlib.iit.ac.lk/xmlui/handle/123456789/211 2017-01-01T00:00:00Z http://dlib.iit.ac.lk/xmlui/handle/123456789/210 TDLang - Type - driven language dependently typed object - oriented JVM programming language for type - driven development Perera, A.D Compilers are edging towards more advanced compilation techniques, a huge factor limiting advances in compilers is the type system of programming languages. Within the last five years the integration of dependent type system into the strong type system in object-oriented programming languages were identified as a new improvement for compilers, and formal verification for building quality software with less amount of unknown bugs. TDLang is a cross-platform dependently-typed object-oriented JVM language for type-driven development with many benefits such as an easy-to-use syntax sugar for object-oriented programmers, auto-generation of validation code and custom exceptions, JVM language interoperability, etc. Sun, 01 Jan 2017 00:00:00 GMT http://dlib.iit.ac.lk/xmlui/handle/123456789/210 2017-01-01T00:00:00Z http://dlib.iit.ac.lk/xmlui/handle/123456789/209 eyeBot - A virtual assistant for customer care solutions Wijesinghe, Y. R. Customer satisfaction has become a vital component of any organization where serving customers as they desire becomes the first step in making new businesses succeed. Interaction with the customers to clarify their requirements can be done in many ways. Some of the methods can be real time conversations by meeting the clients, via voice calls and emails. These mentioned methods consume time and sometimes might not be very responsive which may cause incidents where customers can be missed. With the rapid growth of innovative technologies and social networking methods new businesses tend to use communication methods such as WhatsApp, Viber, Telegram and email groups. As these social networking methods cannot be integrated with the official sites the organizations needs to duplicate and invest time and effort in replicating the answers provided to the customers. Also, only Telegram is compatible with bot integration compatibility where else WhatsApp and Viber does not support these features. Further to maintain these groups there should a dedicated subject matter expert to answer the client questions. With chat bot solutions, the subject matter experts can be eliminated or the questions answered by the subject matter experts can be reduced. With comparison to the traditional methods the SME (Subject Matter Expert) will not have to answer the same question twice as the AI (Artificial Intelligence) chat bot will train itself and answer the questions accordingly. The dissertation discusses this approach to provide a smart chat bot solution to provide an optimized method of providing real-time customer care. Sun, 01 Jan 2017 00:00:00 GMT http://dlib.iit.ac.lk/xmlui/handle/123456789/209 2017-01-01T00:00:00Z