Browsing by Author "Karunaratne, PM"

Now showing 1 - 12 of 12

item: Conference-Full-text
Analyzing the Healthiness of an IT Project Using Self Organizing Maps
Thilakarathne, HHTD; Wellage, CH; Rupasinghe, JAPNS; Nimantha, KC; Karunaratne, PM; Fernando, S
This research paper discusses a solution for a problem we have identified that IT companies face when managing their projects. Project managers often find themselves in a tough situation when deciding the current status of a project and making decisions based on the evaluation. But similar situations have happened earlier in other projects and the knowledge about the measures that were taken at those situations and their effect on the success of the project can be used to evaluate similar situations in new projects. Our approach in this regard is analyzing the past data of IT projects using different machine learning techniques to identify the major factors that have affected the success of a project, understand how strongly each factor is bound to success and then training a model with the data. Where it can be used to analyze situations that arise in new projects and identify how like the current situation is to lead the project in to a success or a failure. The machine learning technique we have likely used in this study is Self- Organizing Maps (SOM) and the system was implemented using Python language
item: Article-Abstract
A Comprehensive information extraction module for reducing call handling time in a contact centre
Gunathunga, KIH; Priyadarshana, YHPP; Perera, KKANN; Ranathunga, L; Karunaratne, PM; Thanthriwatta, TM
Information extraction plays an important role in text related research and application areas such as text mining and dialogue systems. Information extraction can be done using key word extraction and measuring the semantic similarity between texts. These concepts are applied to address a key issue in the telecommunication contact centre domain where the customer dissatisfaction is increasing due to higher call handling time. The proposed method is a combined with a key word based approach and a semantic similarity based approach with the use of semantic nets. The semantic similarity of two sentences is calculated using word similarity and the word order. Experiments on two sets of sentence pairs illustrates that proposed method provides a similar measure which is significantly correlated to human intuition. The overall accuracy of the information extraction module is approximately 70% based on the evaluation results.
item: Conference-Extended-Abstract
Decision supportive system for customers and merchants with enhanced relevance
(2014-01-16) Bandara, YMJK; Madhumalie, SHP; Madurasinghe, SY; Weerasinghe, MSC; Karunaratne, PM
Abstract - The rrsmg development of technologies caused consumers to deal with an increasingly large amount of product and service information which is far beyond the individual's cognitive effort to process. However due to the popularity of ecommerce the number of online retailers proliferated. As a result of this, currently there are large numbers of brand names available online and this results in customer difficulty of finding the best product among alternatives. Since different customers have different perspectives and parameters about products it is better to facilitate customized information. Merchants have problems in improving their businesses due to lack of information about the customer buying pattern variations. Therefore, a system which provides accurate information and suggestions which significantly decrease users' cognitive effort which would be beneficial for both customers and merchants when selecting and providing products and services. "Decision Supportive System for Customers and Merchants with Enhanced Relevance - EpsiShoppingBot" is a mobile and web based solution which addresses above mentioned requirements of both customers and merchants.
item: Conference-Full-text
Decision Supportive System for Customers and Merchants with Enhanced Relevance
Bandara, YMJK; Madhumalie, SHP; Madurasinghe, SY; Weerasinghe, MSC; Karunaratne, PM
The rising development of technologies caused consumers to deal with an increasingly large amount of product and service information which is far beyond the individual’s cognitive effort to process. However due to the popularity of ecommerce the number of online retailers proliferated. As a result of this, currently there are large numbers of brand names available online and this results in customer difficulty of finding the best product among alternatives. Since different customers have different perspectives and parameters about products it is better to facilitate customized information. Merchants have problems in improving their businesses due to lack of information about the customer buying pattern variations. Therefore, a system which provides accurate information and suggestions which significantly decrease users’ cognitive effort which would be beneficial for both customers and merchants when selecting and providing products and services. “Decision Supportive System for Customers and Merchants with Enhanced Relevance - EpsiShoppingBot” is a mobile and web based solution which addresses above mentioned requirements of both customers and merchants.
item: Conference-Extended-Abstract
An effective approach for personalized web search based on community-cluster analysis
(2014-01-16) Thanthriwatta, TM; Karunaratne, PM
The concept of Personalized Web Search is commonly used for improving the quality of web search results by identifying and facilitating different users' search needs. There are several techniques such as user profiling, content analysis, hyperlink analysis and biased PageRank algorithm that are used to achieve web personalization. User Profiling is one of the widely used techniques for personalizing web search at large scale. But it contains several technical and ethical issues such as privacy violations, inefficient use of computing resources as well. Collaborative web search is also a kind of a relatively "new concept which defines the way of optimizing/personalizing search results by using details of group of people and contributing the knowledge of all of them about web search. This paper presents the details of an alternative approach for personalizing web results by using user profiling technique with community cluster analysis of collaborative web search by adapting concept of reusability 'among web results.
item: Conference-Full-text
An Effective Approach for Personalized Web Search based on Community-Cluster Analysis
Thanthriwatta, TM; Karunaratne, PM
The concept of Personalized Web Search is commonly used for improving the quality of web search results by identifying and facilitating different users’ search needs. There are several techniques such as user profiling, content analysis, hyperlink analysis and biased PageRank algorithm that are used to achieve web personalization. User Profiling is one of the widely used techniques for personalizing web search at large scale. But it contains several technical and ethical issues such as privacy violations, inefficient use of computing resources as well. Collaborative web search is also a kind of a relatively new concept which defines the way of optimizing/personalizing search results by using details of group of people and contributing the knowledge of all of them about web search. This paper presents the details of an alternative approach for personalizing web results by using user profiling technique with community cluster analysis of collaborative web search by adapting concept of reusability among web results.
item: Article-Abstract
An ICT Enhanced Life Quality for the Elderly in Developing Countries: Analysis Study Applied to Sri Lanka
Karunaratne, PM; Firdhous, M
In the recent years, the entire world has seen a tremendous increase in the elderly population. Even though countries differ in the numerical criterion for defining the old age, the UN has agreed that the cutoff of 60+ years refers to the older population. When the percentage of older population increases in a country, the country faces new challenges in the form of economical as well as social impacts. The economic impacts are felt in the form of a shortfall in labor supply, income, household savings while an increase in the payment of retirement benefits and healthcare expenditures. On the social side, the elderly population feels continuously isolated due to the changes in the value system. The United Nations has identified five quality of life characteristics for the elderly in its 1991 resolution named the United Nations Principles for Older Persons. In this paper the authors initially look at the trends in the increase of aged population in Asia in general and Sri Lanka in particular. Then they discuss how the quality of life characteristics can be achieved in a cost effective way through the use of Information and Communication Technology (ICT). The authors take an in depth look at the recent developments in field of ICT and its penetration in developing countries especially Sri Lanka with special emphasis to constraints and challenges in adopting ICT for the improving the quality of life of elderly.
item: Conference-Full-text
Iot-powered sustainable dry zone agriculture: an experimental implementation
(Information Technology Research Unit, Faculty of Information Technology, University of Moratuwa, Sri Lanka, 2018) Firdhous, MFM; Sudantha, BH; Karunaratne, PM; Wijesiriwardana, CP
Dry zone agriculture plays an important role in the national economies of many countries. On the other hand dry zone agriculture suffers from many problems especially lack of water for producing the optimum yield it should. Modern technologies such as Internet of Things can be adapted successfully to improve the yield by managing the inputs at optimum levels. This paper presents an experimental IoT powered microclimate management system that is capable of continuously monitoring the important environmental parameters in specific areas. The proposed system not only monitors the soil moisture levels, but also controls the pot irrigation system for managing the water supply at required levels enabling sustainable agriculture. The built-in communication module alerts the parties concerned when the moisture levels falls below the pre-decided levels requiring human intervention.
item: Thesis-Abstract
Online career information display & students information management system for COTs
(2014-08-18) Jayasekara, LUL; Karunaratne, PM
This project is concerned with the development of a system for easy handling of online information & student’s information management. Its main functions are online display of available courses, online applicant’s registration, student’s registration, internal exam candidate’s registration, automated results generation, giving different privileges for different user levels, online results as well as many more students’ management functionalities. This new system will enable automation of all functions, thus conceiving an efficient, reliable speedy and cost effective operational model for the online career information & applicant’s management system for 9 colleges of Technology Island wide. The scope of the project is envisaged that the project is in complete functional compliance to the presently manually performed functions and also to provide for recommended future expansions. Further, this system can be directly implemented to be a more complete (integrated) product by providing access facilities through the web. Target group of members for the system are applicants, registrar, evaluation officer, other colleges and any other general users. Inputs for the system are problems the problems in existing system. Process is the technology used to solve the above problems. Output is the solutions and value added features providing by the new implemented system. For facts finding purposes used formal methods as well as informal methods. When comparing to the existing system which is maintained in manual, with the introduction of the new system, there are useful facilities available. This System is a user friendly, self contained product and can be operated on its own. In brief, the above facts namely, on line display courses & other search details, manage the applications, online applicants registration, students registration, internal examinations management, automated results preparation can be identified as most important features of the new system.
item: Conference-Abstract
Sentiment analysis: Measuring sentiment strength of call centre conversations
Priyadarshana, YHPP; Gunathunga, KIH; Perera, KKANN; Ranathunga, L; Karunaratne, PM; Thanthriwatta, TM
Measuring sentiment strength can be considered as one of the key areas of sentiment analysis. The existing sentiment word scoring functions are based on the intensity of adjectives in the sentences. To date, there is a very minimal work has been done for measuring sentiment strength based on adverb and adjective combinations. This research proposes a sentiment word scoring method considering the adverb-adjective combinations of a given sentence. This is subject oriented hence this approach is proposed for a call centre domain. We use the linguistic analysis of adverbs of degree to satisfy the adverb-adjective scoring. Then a new approach has been introduced to calculate the sentence wise sentiment score by enhancing the existing sentence level sentiment scoring algorithm. Authors propose the way of determining the total sentiment strength of a given sentence. We describe the results of our experiments on set of 100 call centre conversation audio files between the call centre agent and customer and compare our algorithms with the existing sentiment scoring algorithms. It can be concluded that these enhancements have an accuracy of 74% in precision.
item: Conference-Abstract
Sentiment negation: A novel approach in measuring negation score
Priyadarshana, Y; Ranathunga, L; Karunaratne, PM
Sentiment negation and negation scoring can be considered as major aspects of sentiment analysis. Social media sentiment analysis can be considered as an excellent source of information in today's business. But there is very minimal work has been done in sentiment negation scoring. All the existing negation scoring mechanisms are based on an adjective intensity approach. This research proposes a novel mechanism for negation scoring based on an adjective-adverb combined approach. The domain is movie reviews which are obtained from international movie database. We have introduced new negation axioms that all negation scoring mechanisms must satisfy. Three new negation scoring functions which satisfy those negation axioms have been introduced here to calculate word, sentence and document level sentiment negation. These are all based on adjective-adverb hybrid approach. We describe our experimental results on different 700 training and different 300 testing movie reviews. We compare our functions with existing sentiment negation functions. It can be seen that our functions perform better than the existing functions based on Pearson correlation with respect to human subjects. It can be concluded that these new functions perform an overall accuracy of 71% in precision.
item: Conference-Abstract
Synchronized medical image viewer for remote consultation
Karunaratne, PM; Jayasuriya, DPN; Karunaratne, MBRD; Karunasekera, HHR; Nanayakkara, ND
The collaborative analysis between radiologists and surgeons of digitized medical images in the surgical planning phase is a critical step in the process of patient treatment. The shortage of radiologists who are qualified to carry out this task has proven detrimental to patient care. The proposed solution is a tool for collaboration where a radiologist and a surgeon are able to analyze scans while being in geographically disparate locations. Though a superficial examination of the options currently available would suggest the use of an off-the-shelf desktop sharing software solution, these fall well short of the performance and quality requirements of such an application which is to be used in a medical setting. We propose a novel approach, based on low-level command capturing that enables near-real-time synchronization with zero loss of information or distortion of the images. We discuss the underlying technology of our implementation of the synchronization solution as well as present the manifest performance gains of our solution in comparison with the highly popular commercial desktop sharing solution TeamViewer.