Margaret Bernard

Creating sharable learning objects from existing digital course content

Our research is targeting Instructors that have course material as a collection of various digital documents (raw content) and whose objective is to re-structure this raw content into a standards-based format in order to support a higher degree of content reuse, sharing and easier maintenance. In previous work, we differentiated a Reusable Learning Object (RLO) from a Sharable Learning Object (SLO) and developed a model which can be applied to convert RLOs into SLOs [4]. In this paper, we present an iterative five-step method to re-structure selected raw content into RLOs. The model from the previous work is then applied to convert the RLOs into SLOs. Thus far, we have used raw content from one Instructors Computer Architecture course and found that conversion of the raw content can successfully result in a subset of the raw content residing in SLOs, a form which is more conducive to reuse, sharing and content maintenance. In ongoing work, we are applying the methodology to additional raw content from several other Instructors (Computer Science courses) with a view to refining and automating the process where possible.

Intelligent diagnosis of diseases in plants using a hybrid Multi-Criteria decision making technique

An Expert System that can intelligently diagnose diseases in plants is proposed.A model for framing diseases and their pathological characteristics is presented.The disease diagnosis method uses Multi-Criteria Decision Making techniques.The AgriDiagnose system consists of a Pathology tool and a mobile app.Experimental results show an accuracy of over 95%. This paper describes an Expert System that can intelligently diagnose diseases in plants. The system is dialog-based and uses a Multi-Criteria Decision Making technique that is a hybrid of Analytic Hierarchy Process and Sensitive Simple Additive Weighting. The paper describes an approach for disease modeling that uses a set of characteristics which are weighted for each disease using two types of weights: Relative Weights and Scales. The diagnostic process involves calculating the utility value for each disease based on the utility values of its characteristics. Experimental results show an accuracy of over 95%. The system implemented is called AgriDiagnose and it consists of a web-based pathology tool to model the diseases and a mobile app for farmers to interact with the system for disease diagnosis in the field.

A lower bound on average codeword length of variable length error-correcting codes

Codes of variable length that are capable of error correction are studied. A lower bound on the average length of such codes under the criterion of promptness is obtained. This generalizes the celebrated result due to C.E. Shannon (1948) for the noiseless case. The bound obtained provides a measure of optimality for variable-length, error-correcting codes. >

Enhancing the Metacognitive Skill of Novice Programmers Through Collaborative Learning

Computer Supported Collaborative Learning (CSCL) aims to improve education by combining collaborative learning with modern information and communication technology. The opportunity exists to develop successful CSCL applications due to the increase in popularity of social networking and online gaming among students. In this chapter, we present an approach for promoting metacognition in computer programming using collaboration and computer games. We show that CSCL can improve the students’ metacognitive skills and the use of games motivates and engages students in the learning process. Together, they enhance the qualities of a successful problem solver and low problem solving skill has been identified as the main challenge faced by novice computer programmers.

Variable Length Perfect Codes

Abstract In this Daper, variable-length perfect codes are introduced. These are defined as codes satisfying a combinatorial result (published in an earlier paper by authors) which is a generalization of the Hamming sphere packing bound. An e-error correcting, variable-length, perfect code can correct up to e-errors in each segment of the codewords and no more. The paper gives a construction method for some classes of variable length perfect codes.

AgriMaps: Improving site-specific land management through mobile maps

Abstract Population pressures have led to crop cultivation taking place on steep slopes and other marginal lands in the small island developing states of the Caribbean. This threatens regional food security and can result in environmental degradation if site-specific land management practices are not adopted. AgriMaps is a mobile application developed to support decision making in the agricultural sector of Trinidad and Tobago. This application uses an evidence-based approach to making crop and mitigative land management recommendations. The app is a platform for spatial data visualization and provides a greater range of geo-spatial information than is currently available to stakeholders in the farming community. It is envisioned that this application will improve land use planning decisions and the adoption of sustainable agricultural management techniques. We provide the theoretical and technical frameworks needed to develop similar ICT tools in other territories.

An Incremental Editor for Dynamic Hierarchical Drawing of Trees

We present an incremental tree editor based on algorithms for manipulating shape functions. The tree layout is hierarchical, left-to-right. Nodes of variable size and shape are supported. The paper presents algorithms for basic tree editing operations, including cut and paste. The layout algorithm for positioning child-subtrees rooted at a given parent is incrementally recomputed with each edit operation; it attempts to conserve the total display area allocated to child-subtrees while preserving the user’s mental map. The runtime and space efficiency is good as a result of exploiting a specially designed Shape abstraction for encoding and manipulating the geometric boundaries of subtrees as monotonic step functions to determine their best placement. All tree operations, including loading, saving trees to files, and incremental cut and paste, are worst case O(N) in time, but typically cut and paste are O(log(N)2), where N is the number of nodes.

Linear Codes with Non-Uniform Error Correction Capability

This paper introduces a class of linear codes which are non-uniform error correcting, i.e. they have the capability of correcting different errors in different code words. A technique for specifying error characteristics in terms of algebraic inequalities, rather than the traditional spheres of radius e, is used. A construction is given for deriving these codes from known linear block codes. This is accomplished by a new method called parity sectioned reduction. In this method, the parity check matrix of a uniform error correcting linear code is reduced by dropping some rows and columns and the error range inequalities are modified.

Combining eLearning and P2P File-Sharing Technologies

The existing set of Peer-to-Peer (P2P) digital repositories were not originally and purposely designed for sharing learning content, and as such are not well suited in the same way that centralized repositories are for the storing and retrieving of learning content for instructional use. This survey paper therefore reviews some P2P file-sharing and related eLearning technologies, to identify core reasons why there are no P2P eLearning Systems with features comparable to the ubiquitous client/server eLearning Systems. It concludes that P2P sharing of Reusable Learning Content requires a redesign in order to facilitate more comprehensive P2P eLearning Systems.

Peer-Review of Reusable Learning Objects for eBooks in a Peer-To-Peer Network Architecture

It is a challenging task to take standards-based eLearning content from a client-server environment and place it on a Peer-To-Peer network to achieve similar goals. The client-server eLearning system architecture is designed for storing, sharing and delivering course content. In an initial pursuit of this challenge, a design for a Peer-ToPeer Reusable Learning Object Repository was proposed and prototyped. Now in this continuing work, the challenge of performing and managing peer-reviews for the content being shared on the Peer-To-Peer network is investigated, along with how to aggregate and sequence the content into a course for delivery. The solution introduces a scalable server design, mainly to maintain the integrity of peer-review data while leaving the eLearning content on network peers. From the prototype, the user can also now aggregate and sequence peer-reviewed Reusable Learning Objects into a course after which it is passed to a commercial compiler via scripting, along with preprogramed templates to produce eBooks. The result, is that this extended Peer-To-Peer prototype application demonstrates an end-to-end solution which can be used to manage and share peer-reviewed Reusable Learning Objects over a Peer-To-Peer network, and achieve the final goal of packaging and delivering shared content as online and offline eBooks.