http://dlib.iit.ac.lk/xmlui/handle/123456789/1089 2026-04-21T07:09:26Z 2026-04-21T07:09:26Z Fernando, Shanuka http://dlib.iit.ac.lk/xmlui/handle/123456789/1260 2023-01-04T04:12:13Z 2022-01-01T00:00:00Z

Classifying Twitter Posts for Movie Success Prediction using a BERT based Approach Fernando, Shanuka "Predicting a movie's success is a difficult undertaking because the movie industry is expanding exponentially. Many stakeholders rely on the movie's revenue, and marketing techniques can be affected by the outcome of forecasts to increase audience reach and increase revenue. Sentiment analysis, a branch of natural language processing (NLP) that employs NLP methods to extract emotions from text, has been used in several studies to forecast movie success. Several studies have employed sentiment analysis, a subset of natural language processing (NLP) that employs NLP techniques to extract emotions from text, to predict the box office success of films. The majority of studies that used sentiment analysis to predict the performance of movies used conventional research techniques to conduct their research. It has been noted that recent developments in sentiment analysis, such as transformer-based language models, which have the power to drastically alter outcomes in other domains, have not yet been introduced and applied to the domain of predicting movie success. Only a few attempts have been made to increase sentiment analysis's accuracy level. As a result, this study used Twitter posts to introduce a new transformer-based sentiment analysis model to the domain of predicting movie success. With an F1 Score of 89.68 in multi-class classification, a multi-class categorization of tweets as negative, neutral, and positive was able to outperform the existing approaches."

2022-01-01T00:00:00Z Bandara, Thisun http://dlib.iit.ac.lk/xmlui/handle/123456789/1259 2023-01-04T04:05:59Z 2022-01-01T00:00:00Z

Analysing Price Volatility for Price Forecasting of Food Commodities in Sri Lanka Bandara, Thisun There are several types of food commodities available in the Sri Lankan market and the customers usually visit physically or use online services to purchase those items for the fulfilment of their food consumption. The price of those food commodities is one of the main concerns and the sudden price gains and reductions are concerned by several parties who are involved in the buying and selling process. The final outcome of this research project is a solution for the problem of price volatility of food commodities in Sri Lankan market due to the unavailability of a system to deliver price forecast to get an idea about them. LSTM deep learning model approach is identified by analyzing the previous research work, currently implemented solutions and techniques that will result on building up a price forecasting system which comprises with a generalized model. It will deliver the price forecast for the beneficiaries in a user friendly way by maintaining 70% of test accuracy. By viewing those forecasts, the users will be able to have an idea on how the price will be evolved for those commodities in future.

2022-01-01T00:00:00Z Kodippily, Rajeev http://dlib.iit.ac.lk/xmlui/handle/123456789/1258 2023-01-04T03:55:03Z 2022-01-01T00:00:00Z

Handwritten Source Code Recognition For Python Kodippily, Rajeev "Modern programming Integrated Development Environments (IDE's) use keyboard and mouse as their main input method for text entry. However, some programmers suffer from disabilities such as Repeated Strain Injury (RSI) which cannot be productive using traditional text entry methods such as keyboard and mouse. This project focuses on creating a programming IDE in which handwriting would be the main input method for text entry. For this, the research involves training models using the open source OCR Tesseract and comparing the results to a state of the art English handwriting recognition engine in the Google ML Kit. The test results shows that the trained model performs better than base Tesseract on source code with 11.82% less Character Errors and 12.51% less Word Errors. Furthermore the ML Kit engine outperforms the trained model with 16.4 % less Character Errors and 26.5 % less Word Errors resulting in a 12.02 Character Error Rate and 41.65 Word Error Rate. Comparing this result to the benchmark presented turned out to be problematic due to the inaccuracy of the approach and is discussed in the results. The project has the potential to be beneficial for software engineering professionals, educators and students who write computer programs in their day-to-day life and who are looking for alternate input methods to the traditional keyboard and mouse. "

2022-01-01T00:00:00Z Amarasena, Harindu http://dlib.iit.ac.lk/xmlui/handle/123456789/1257 2023-01-04T03:50:08Z 2022-01-01T00:00:00Z

CSLS (CEFR Based Lexical Simplification) Amarasena, Harindu Common European Framework of Reference for Languages (CEFR) is an international standard for describing language ability. There are six levels in the CEFR and most of the English text in the world can be classified into on on these six levels. And everyone who knows English in the world can be classified into one of those levels as well. So there can be a scenario where the persons level and the level of the text that they are reading might be mismatched and and the person might have a harder time understanding what they are reading. So the solution that is mentioned in the paper wishes to solve this issue by creating a system where the CEFR level of a given text is translated in to a higher or a lower level making it easier for the reader to understand it. The developed system works by taking a paragraph that the user enters, calculates its CEFR level, takes a CEFR level from the user and translates the paragraph to that level without changing the original meaning.

2022-01-01T00:00:00Z