http://dlib.iit.ac.lk/xmlui/handle/123456789/3 2026-04-08T12:59:33Z http://dlib.iit.ac.lk/xmlui/handle/123456789/2295 Travel trend analyzing and intelligent tour route recommendation system Weerakkody, Madhawa Tourism is a fast-growing industry and worldwide, tourism rebounded strongly with the growth of online applications on travel and hospitality. This transformation leads to millions of usergenerated online contents on various travel-related digital channels and platforms. Majority of travelers use the internet as an information source for planning trips. Travelers read other travelers' online reviews to narrow down choices. The unstructured text nature makes it harder to understand the idea behind the review and the biased state of mind of the review, hard to derive valuable insights from reviews, makes the processing review contents more difficult and may leads the user to an unambiguous direction. In present-world all are having a value for a day. If travelers searching for reputable sources, sifting through online reviews and doing research as a holiday planning before happening it will be a big challenge as well as it will be a tedious and a time-consuming task. The goal of the research presented was to explore how other travelers’ reviews are utilized in trip planning process and what the observed trends of those travel destinations in order to overcome the above-mentioned obstacles more efficiently by reducing number of human hours spent on review analyzing. The solution is to provide a month based trend analysis of traveling destinations as well as an intelligence travel route planner with recommendations by analyzing reviews and rating in travel review sites as identified throughout the requirement gathering phase by involving different elicitation techniques. The solution is implemented using Natural Language Processing and Text pre-processing approaches under machine learning techniques are playing a great role in this research in order to carry out the sentiment analysis and word classification modules. And performance testing process was carried out considering the accuracy and the efficiency of the system. The testing approach carried out Software functional quality testing and Software structural quality testing. The evaluated results reviled that the accuracy of the sentiment analysis and word clarification modules are in the satisfactory level. A critical evaluation process was carried out based on the different evaluation criteria, involving various evaluator groups. The results of the evaluation process stated the strengths and limitation of the project and several enhancements were suggested. 2018-01-01T00:00:00Z http://dlib.iit.ac.lk/xmlui/handle/123456789/179 “Ankler” Treatment plan application for Ankle Sprain Haputhanthri, Chamika T. The Project mainly focused on Business process models and notations converting into XML format. Business process models are very popular in designing the process models now days. But this research mainly focusses on finding the image processing approach to convert business process model into resource description framework. Also, conversion should be done without any knowledge lost. It is very difficult to find a proper way to convert diagram into computer readable language. Because, recognizing the relationship between the notations and connect the path should be done in proper manner. This project mainly focuses on Appling image processing to recognize relationships and the nodes of diagrams. 2018-01-01T00:00:00Z http://dlib.iit.ac.lk/xmlui/handle/123456789/178 Stock Market Price Prediction using ARIMA Barthelot, Steve Adrian Predicting the behaviour of stocks is not a simple task. With millions of stocks trading every day predicting the price of a stock is a lucrative trade. This project tries to predict the price of a stock as time goes on using ARIMA model in data mining 2018-01-01T00:00:00Z http://dlib.iit.ac.lk/xmlui/handle/123456789/177 Fraudulent Transaction Detection in Bitcoin Network Based on Unsupervised Learning Pathiraja, Fajita Yasas Financial frauds usually means, changing the ownership of property through illegal ways. Fraud is a crime. In financial world frauds take place due to many reasons. Bitcoin is a cryptocurrency (Digital currency) based on blockchain technology introduced in 2009. Frauds are common in bitcoin transactions as well as in traditional transaction methods. Bitcoin transaction process is very different from traditional systems. Due to the pseudonymous behavior of bitcoin reported/identified frauds are very low. This project aims to identify potential frauds by analyzing the blockchain transaction dataset with the help of anomaly detection and unsupervised learning methods. Some of the known cases detected throughout this analysis. 2018-01-01T00:00:00Z