Madhushi Bandara

A Complex Event Processing Toolkit for Detecting Technical Chart Patterns

With the advent of large high volume data, we have seen need for real time analytic techniques like Complex Event Processing. This paper extends a Complex Event Processing Engine to support real time identification of technical chart patterns from streaming data. Technical chart patterns are known interesting recurring patterns on time series data, and they are used by experts in time series data analysis domains such as stock market and currency exchange rates. Yet the automated identification of these patterns is challenging due to the high volatility and noise of data. The paper focuses on identifying suitable technique to filter out volatility and a set of algorithms to query the data streams continuously and to identify patterns. The resulting solution is a toolkit for chart pattern recognition which is a composition of a set of complex CEP queries and a Kernel regression smoother applied on moving windows. Same toolkit can be used to detect chart patterns in other domains like Gold and Oil prices etc as well.

A traceability management framework for artefacts in self-adaptive systems

We expect software systems to be dependable and sufficiently responsive to the inevitable changes regularly happen in their operational environments. This can be a challenging task to achieve when systems are in enterprise scale and large enough to cater for multiple complex business processes. One approach to address this is by incorporating suitable software process models and managing various artefacts within the process. However, once deployed, only maintenance of the software is viable through the process; it may not be sufficient for the needed changes due to the essential difficulties associated with software engineering. In order to overcome this challenge, self-adaptive systems with dynamically modifying architectures are becoming popular and sufficiently warrant for a mainstream practice in future. In light of these developments, we identified an important yet missing part in self-adaptive system engineering, i.e. artefact traceability management at runtime. This paper presents the research work on developing a generic traceability management toolkit with a traceability interlink visualizer aiding software engineers to explore artefact inconsistencies rapidly. The toolkit was then extended with a widely used self-adaptive system framework. The evaluation of the developed traceability management framework with a case system is presented.

Evaluating the impact of DevOps practice in Sri Lankan software development organizations

The purpose of this study is to conduct an analysis on practicing DevOps in software development companies in Sri Lanka. DevOps is extended from agile with a mix of patterns intended to improve collaboration between development and operations. The main objective of this research is to identify whether there is a relationship between quality, responsiveness to business needs and agility with implementation of DevOps. Other objectives of this research are to identify barriers Sri Lankan software companies have faced and provide recommendations to overcome those issues. This study was carried out using a deductive approach. Data have been collected from IT professionals in Sri Lanka who had prior experience or knowledge about DevOps. Recommendations are given based on interview feedback, hypotheses output and descriptive statistical analysis. According to results software professionals in Sri Lanka are in a perception that implementation of DevOps has a positive impact on Quality, Responsiveness to Business Needs and Agility to New Technologies. Culture, Automation, Measurement and Sharing are important factors to be considered when implement DevOps in Sri Lanka. As a result of this study it is recommended to implement and practice DevOps in Sri Lankan Software companies.

Big Data Analytics Has Little to Do with Analytics

As big data analytics is adapted across multitude of domains and applications there is a need for new platforms and architectures that support analytic solution engineering as a lean and iterative process. In this paper we discuss how different software development processes can be adapted to data analytic process engineering, incorporating service oriented architecture, scientific workflows, model driven engineering and semantic technology. Based on the experience obtained through ADAGE framework [1] and the findings of the survey on how semantic modeling is used for data analytic solution engineering [6], we propose two research directions - big data analytic development lifecycle and data analytic knowledge management for lean and flexible data analytic platforms.

Semantic Model Based Approach for Knowledge Intensive Processes

Many business processes present in modern enterprises are loosely defined, highly interactive, involve frequent human interventions. They are coupled with a multitude of abstract entities defined within an enterprise architecture. Further, they demand agility and responsiveness to address the frequently changing business requirements. Traditional process modelling and knowledge management technologies are not adequate to represent and support those processes. In this paper, we discuss how a process management system based on semantic models can be used to address the needs of non-traditional and knowledge intensive processes. The modelling capabilities of the framework are demonstrated via a case study and evaluated using set requirements that KIP supporting process management system should have. Finally, we discuss how this semantic model based solution can be improved further to cater for the management and execution of knowledge-intensive business processes in a broader context.

Visualising complex event hierarchies using relevant domain ontologies: Doctoral Symposium

With the growth of data available for analysis, people in many sectors are looking for tools to assist them in collating and visualising patterns in that data. We have developed an event based visualisation system which provides an interactive interface for experts to filter and analyse data. We show that by thinking in terms of events, event hierarchies, and domain ontologies, that we can provide unique results that display patterns within the data being investigated. The proposed system uses a combination of Complex Event Processing (CEP) concepts and domain knowledge via RDF based ontologies. In this case we combine an event model and domain model based on the Financial Industry Business Ontology (FIBO) and conduct experiments on financial data. Our experiments show that, by thinking in terms of event hierarchies, and pre-existing domain ontologies, that certain new relationships between events are more easily discovered.

An automatic classifier for exam questions in Engineering: A process for Bloom's taxonomy

Assessment is an essential activity to achieve the objective of the course being taught and to improve the teaching and learning process. There are several educational taxonomies that can be used to assess the efficacy of assessment in engineering learning by aligning the assessment tasks in line with the intended learning outcomes and teaching and learning activities. This research is focused on using a learning taxonomy that fits well for computer science and engineering to categorize and assign weights to exam questions according to the taxonomy levels. Existing Natural Language Processing (NLP) techniques, Wordnet similarity algorithms with NLTK and Wordnet package were used and a new set of rules were developed to identify the category and the weight for each exam question according to Blooms taxonomy. Using the result the evaluators can analyze and design the question papers to measure the student knowledge from various aspects and levels. Prior evaluation was conducted to identify most suitable NLP preprocessing techniques to the context. A sample set of end semester examination questions of the Department of Computer science and Engineering (CSE), University of Moratuwa was used to evaluate the accuracy of the question classification; weight assignment and the main category assignment were validated against the manual classification by a domain expert. The outcome of classification is a set of weights assigned under each taxonomy category, indicating the likelihood of a question to fall into a certain category. The highest weight category was considered as the main category of the exam question. According to the generated rule set the accuracy of detecting the correct main category of a question is 82%.