Private study with search engine support - Examination papers

Abstract

There is a growing tendency for model answers to past examination questions to be made available to students. Laudable though this may be, it does encourage some to concentrate on how to answer the usual questions that occur every year, rather than to learn the whole content of the module. Those students may achieve higher marks than they would otherwise do, but would they be less able to apply the content in novel situations? This paper records an evaluation of a proposed system that allows individual parts of past examination papers to be submitted to a search engine. The files used by the search engine are derived from a series of narrated and animated online lectures in digital circuit design. The resulting returns purport to offer a recommendation as to those lecture slides that are relevant for answering the question. The paper briefly describes the essential characteristics of the on-line lectures and how they are specified. The search mechanism is based on a conventional term frequency-inverse document frequency scheme. The post processing of the results prior to display are described and a mock-up of the interface is illustrated. The paper concludes with an objective and discursive evaluation of the results and indicates the path for future work.