PURSE 2005

Permanent URI for this collection

Browse

Recent Submissions

Now showing 1 - 5 of 9
  • Item
    Applying object role modeling to software design patterns
    (University of Peradeniya, Sri Lanka, 2005-11-10) Karunaratne, H. T.; Kodithuwakku, S. R.
    The most important issue in object-oriented design is identifying classes and their relations in the problem domain. Relations such as association and. inheritance should be specified clearly. But problems arise when class migration comes into account in the problem domain. In such cases, the user has to consider several classes at the same time. Researchers have identified that incorporating roles into object-oriented design can solve this problem. This approach is known as Object Role Modeling (ORM). The Object Role Modeling methodology developed by Trygve Reenskaug presents a general approach to modeling objects and object collaborations using roles and role models. The aim of this research is to apply ORM to reduce complexity when doing 'large scale' design. This is done by supporting separation of concern, a flexible design approach and reusable designs. At present, developers use 'Design Patterns' to develop reusable software. The main problem is that the patterns have been described using the conventional object model, hence not applicable when 'object reclassification anomaly' is presented in the problem domain. There is a need in describing patterns with alternative ways to address these issues. In this research, we have applied ORM to overcome the problems and limitations of class-based modeling -when designing with patterns. With ORM, designing and using of patterns can be described succinctly and with much better separation of concern than with classes. Design patterns will make the problem more simple and reusable while roles make the design more accurate. Since the ORM concept is becoming more and more popular, describing patterns using 'roles' can be more advantageous. Roles will handle the problem of class migration as well as supporting the separation of concern. Seven commonly used patterns are chosen in this study to redesign using roles. Furthermore, a new role constraint is introduced to easier the combination of these two concepts.
  • Item
    Application of tracer techniques to investigate water leak: a case study from Nortonbridge reservoir
    (University of Peradeniya, 2005-11-10) Jayakody, Jeevan; Gunatilake, Jagath
    The Nortonbridge reservoir was constructed across the 'Kehelgamu Oya' in 1953. Recently, a critical leakage problem was encountered where water emerged out from two nearby springs approximately 50 m downstream of the dam on the right bank of the Kehelgamu Oya. This is not only an economical loss to the project but also a threat to the safety of the abutments, dam and appurtenant structures. Field investigations supported by tracer analysis were used to identify ingress location along the 'Agra Oya.' Subsequent detailed geological mapping carried out in this area could specify four possible ingress locations. Application of tracers such as uranine and odium chloride as well as turbid water indicates that the ingress areas are not within the reservoir but along the Agra Oya, about 450 m upstream from the Nortonbridge dam. Major causative factor for the leakage is relatively high intensity of open joints, which extend from Agra Oya to downstream areas to the Kehelgamu Oya. We suggest construction of an impervious blanket by temporarily diversion of Agra Oya as a remedial measure to seal off the joints.
  • Item
    The Association of the Iorn status with educatioinal performance and intelligence among adolescents in the district of Kandy
    (University of Peradeniya, Sri Lanka, 2005-11-10) Dissanayake, D. S.; Kumarasiri, P. V. R.; Nugegoda, D. B.; Dissanayake, D. M.
    Unavailability of data on the relationship between the iron status and the cognitive functions in a Sri Lankan population led to this study. A cross sectional comparative study was carried out among 13 - 15 year old students selected from 14 schools in the Kandy district. The haemoglobin and the serum ferritin levels were assessed to determine the iron status. The educational performance and the intelligence were compared between iron deficient and iron sufficient groups (n = 188). Each selected iron deficient student was matched with an iron sufficient student from the same school, same class and of same sex. The educational performance was assessed based on the marks obtained for mathematics, science, social science and Sinhala language. The Raven's Standard Progressive Matrices was used to measure the intelligence. An interviewer administered questionnaire was used to collect details on factors that could be effect modifiers or confounders to the relationship of the iron status with educational performance and the intelligence. Serum zinc level and the free thyroxine level were determined in 25 % of the sample. Home visits were carried out in a randomly selected sub-sample to check the quality of data and to observe the horne environment. Neither the educational performance nor the intelligence showed a significant association with the iron status of adolescents in the univariate analysis. However, 23 co- variables and 8 co-variables were significantly associated with the educational performance and the intelligence respectively. The iron status did not show a significant relationship with the educational performance or with the intelligence even when the effects of all these variables were controlled in a multivariate analysis. Following a linear regression analysis; the intelligence, enthusiasm of the student towards learning, ambition, household possessions, problems at home and private tuition for mathematics were the key factors identified as predictors of educational performance. The educational level of the mother and the height for age of the student were the most important factors influencing the intelligence. Further, the home visits revealed that the adolescents experiencing unsatisfactory home conditions had a significantly less educational performance. Zinc deficiency and thyroxine deficiency were commoner among iron deficient students than among iron sufficient students. However, according to the findings, it is unlikely that these deficiencies would have been confounders to the association between the iron status and cognitive functions among these adolescents. According to the findings, the iron status does not play a major role in educational performance and the intelligence of school going adolescents in the district of Kandy. There is no reason why it should not be applicable for school going adolescents in Sri Lanka in general.
  • Item
    An analysis of the well being of Samurdhi beneficiaries
    (University of Peradeniya, 2005-11-10) Gunawardena, A. U. H.; Ariyawardana, A.; Weligamage, P.
    Samurdhi programme is the largest poverty-alleviating programme operating in Sri Lanka. Main aim of the Samurdhi programme is to promote self-reliance by nurturing the savings habit and developing income-generating self-employment. Since development of micro enterprises and dissemination of appropriate technology is regarded as a better way of reducing poverty, government invests on micro enterprise development through the Samurdhi programme. Therefore, this study was designed to compare the well being of the Samurdhi beneficiaries with respect to their business activities. Stratified random sampling was used to select the respondents. Kandy district was selected to conduct the study purposively and of the 120 Samurdhi beneficiaries, 60 involved in income generating activities and 60 who were not involved were selected from three OS divisions: Gampola, Yatinuwara and Kundasale. Data were collected by using a pre-tested questionnaire. Principal Component Analysis was used to construct the household asset related wealth index. Results revealed that 53.3% of the micro enterprise owners were females and 85% of the owners have had the access to credit. Of the group who are involved in micro enterprises, 48.3 % have had formal education up to GCE (OIL). Lack of skills and support from the family members were the main reasons for others not to involve in micro enterprises. Of the group of micro enterprise owners, 30 % were involved in trade related activities like retail shops and doing businesses in the pavement, 18.3 percent in service provision like working in service stations. However, only 3.3 percent were involved in crop based agricultural activities such as mushroom cultivation. The group that was not involved in micro enterprises was mainly engaged in casual work. Respondents were categotised into well-being groups based.on the asset index constructed. The variables namely: availability of electricity, quality of wall material, quality of floor material, quality of roof material, sanitation facilities, value of furniture and value of appliances were used in constructing the asset index. The constructed asset index explained 52.7 percent of the total variance and ranged from - 2.522 to 2.031. Eighty percent of micro enterprise owned households falls within the highest well-being group. The t test confirms that the difference of means when compared between groups is significant. It could be concluded that micro enterprise owners are relatively wealthier than the people who are not involved in micro enterprises. Thus, promoting micro enterprises can be regarded as a better way to reduce poverty among low-income earners.
  • Item
    Affective variables, learning approaches and academic achievements
    (University of Peradeniya, 2005-11-10) Sugathapala, R.D.
    Academic learning outcomes are believed to be influenced I by a complex system that involves the interaction of student personality characteristics, and learning processes. There have been few studies which have examined the interrelationships between these variables and their causal effects on achievement outcomes. This study aims to investigate the interrelationships of affective variables, (self concepts and locus of control) learning approaches, and academic achievement. It was hypothesized that academic locus of control and academic self-concept affected the learning approaches the student adopted and subsequently influenced the achievement outcome of students. The participants were 297 male and female G.C.E. Advanced Level students who were in a l-AB boys school and a girls school located in the Kandy town area. Measures of academic self concepts and locus of control and learning approaches and academic achievement were obtained. Descriptive and inferential statistic methods were used to analyze the data. Structural Equation Modeling (SEM) techniques were used to test the relationship among the variables. Results indicated that both academic locus of control, and academic self concept influenced academic achievement indirectly via students' learning approaches. The findings support the theoretical notion that personality variables influence student learning process and also their academic achievement subsequently.