Visualisation is often presented as a means of simplifying information and helping people understand complex data. With this in mind, Chris Roast has been investigating how visualisation can be used to help users design spreadsheet formulae.
Spreadsheets are a traditional, common and accessible ICT tool that is often used to perform any number of numeric activities where users can quickly become embedded in ‘solutions’ in an information infrastructure that is intrinsically easy to mismanage.
In the development of the formula visualisations, a process of iterative refinement was used, engaging an audience of mid-teen learners and involving re-design and formative evaluation. The resulting visualisation techniques have been found to be broadly relevant to spreadsheet users beyond the initial target audience. The resulting system, EQUS, has since been developed as a fully integrated plug-in for MS Excel. EQUS can be accessed via VeryViz.
Word co-occurrence automated analysis
Laurie Hirsch has been investigating the problem of automated analysis of text documents using word co-occurrence. Words that appear close to each other and appear together a number of times may give an indication of the semantics of the text. A fast and automatic way of measuring this could be useful in search and classification problems.
Laurie has developed a unique approach to co-occurrence analysis in an algorithm that applies a decaying function to the distance between word pairs found in a text such that words regularly occurring close to each other score highly. Words occurring some distance apart will still make a small contribution to the overall co-occurrence score.
This is in contrast to other algorithms which simply count adjacent words or use a sliding window of fixed size. Laurie has created a tool to implement his algorithm which employs a variety of interactive tree and graph visualisations to present results to the user in an easily understandable way.