http://dlib.iit.ac.lk/xmlui/handle/123456789/39 2026-04-06T22:07:46Z 2026-04-06T22:07:46Z Wimalagunaratne, Minul Poravi, Guhanathan http://dlib.iit.ac.lk/xmlui/handle/123456789/450 2020-05-27T18:22:25Z 2018-01-01T00:00:00Z

A Predictive Model for the Global Cryptocurrency Market: A Holistic Approach to Predicting Cryptocurrency Prices Wimalagunaratne, Minul; Poravi, Guhanathan The realm of cryptocurrency has grown exponentially over the past decade, with the most rapid advances seen in the past few years as more and more parties around the world recognize the value of holding digital assets online. Statistics from Twitter support this statement where, approximately 1,500 Tweets about Bitcoin alone is recorded per hour. Consequently, many people are beginning to become more aware and accepting of the nature of digital currencies, and traders in particular seek to know how they can make profitable crypto-coin trades and investments. Although a number of research projects have been undertaken to develop systems that can effectively predict price movements in the cryptocurrency market, they display significant efficiency gaps, which this paper further explores. The authors then attempt to learn from past studies and construct a more holistic approach to a predictive price model for the cryptocurrency market. This focuses on assessing key factors that affect the volatility of the market - public perception, trading data, historic price data, and the interdependencies between Bitcoin and Altcoins - and how they can be best utilized from a technological aspect by applying sentiment analysis and machine learning techniques, to increase the efficiency of the process.

2018-01-01T00:00:00Z Weerakoon, Shayne Poravi, Guhanathan http://dlib.iit.ac.lk/xmlui/handle/123456789/449 2020-05-27T18:18:05Z 2018-01-01T00:00:00Z

Knowledge Extraction from Question and Answer Platforms on the Semantic Web: A Systematic Review of Technologies Available for Information Extraction Weerakoon, Shayne; Poravi, Guhanathan Knowledge Extraction is the process of getting structured data from unstructured or semi-structured sources. Much research has been conducted in this field and applying these technologies to the web has become a key effort in the past few years. This is due to changes from web 1.0 where the web was simply a set of static pages where user interaction was minimal. With the rise of web 2.0, the internet is no longer a medium to access static information. Users can now share their own thoughts easily thus increasing the amount of user generated content. This has made the web ripe with knowledge, however not all information can be easily accessed. This paper aims to bridge the gap between knowledge available and the knowledge accessed using knowledge extraction.

2018-01-01T00:00:00Z Senthurvelautham, Sharanjaa Hettiarachchi, Saman http://dlib.iit.ac.lk/xmlui/handle/123456789/448 2020-05-27T18:12:47Z 2018-01-01T00:00:00Z

SentScore: Autonomous Text Sentiment scoring and Summarizing System related to Complaint Management Senthurvelautham, Sharanjaa; Hettiarachchi, Saman In traditional markets, customer complaints are considered as an important source of information. Since complaint management is recognized as a central for customer satisfaction, any measure of complaint behaviour should consider the degree and quality of the underlying customer satisfaction. Therefore, analyzing customer complaints is part of the process of a business. A prompt, reasonable and efficient response to a complaint can win you a loyal customer, and develop your business's reputation for top quality service. This project would be analysing customer complaints, in order to improve customer experience. As the solution to solve this issue, the proposed solution would address issues with respect to consumer complaint data in a textual format (complaint by phones), which are identified with the IT field (Technical Support Complaints). Furthermore, literary data written in English dialect will be considered. Moreover, SentScore ought to be savvy enough to interpret data identified with complaints efficiently and effectively, classify and analyse sentiment score precisely, summarise them into aspects, and distinguish how the customer feels about those aspects. With this proposed solution the Customer Complaint Operators are able to extract a summarized analysis of the complaint solution by assigning weights to the complaint and aspects including Internet, Television and Facility, which are the main aspect categories considered when analyzing the customer complaint. The system makes utilization of Natural Language Processing, Machine Learning and Sentiment Analysis concepts, to provide the highest accurate sentiments or opinions expressed by the customer in complaints, to present the end users with accurate and effective summarized outcome of the customer complaints and aspect of it.

2018-01-01T00:00:00Z Selvanathan, Nikethan Poravi, Guhanathan http://dlib.iit.ac.lk/xmlui/handle/123456789/447 2020-05-27T17:57:04Z 2018-01-01T00:00:00Z

Comparative Study on Decentralized Cloud Collaboration(DCC) Selvanathan, Nikethan; Poravi, Guhanathan Cloud collaboration is a billion-dollar industry, for sharing, storing, and co-authoring files. In the current age of information technology, cloud collaboration expects to see a significant amount of growth, as more organizations look to leverage the benefits of the industry specifically in the areas of flexibility, cost-efficiency, and security[1]. However, existing systems basically operates in a centralized cluster to achieve high performance, though they have a demand solving indisputable benefits, there are several inherent weaknesses such as high server costs for service providers, illegal data mining in trust-based architecture, security loopholes, and unethical government surveillance. Therefore, a large-scale resource sharing decentralized system can mitigate these traditional server expenses, data failure, and outage, as well as the enhanced security, and privacy of data. This dissertation presents a background to the problem, its impact on adaption, existing research background, and proposing design for storing, sharing, and coauthor files. The Design presents a decentralized resource (storage and network) sharing system, with real-time collaborative editing, peer (node) management, and redundancy schemes to manage fault tolerance of the distributed storage.

2018-01-01T00:00:00Z