Project Proposal by Martin Stacey
Computational Careers Advisor
Software |
OO language such as PHP or JSP, plus HTML and CSS; possibly LISP or Prolog or an expert systems shell |
Covers |
Artificial intelligence, psychology of personality, possibly databases or statistics. |
Skills Required |
Programming, interest in AI, interest in psychology of personality |
Challenge |
Conceptual Technical Programming |
Brief Description
Lots of people choosing degrees, or choosing final year projects, or choosing jobs or postgraduate courses at the end of their degrees, have no idea what sorts of jobs would suit their personalities as well as their abilities. This is partly because they don't know some careers even exist, and partly because they don't know what many jobs really involve: they may be eager to leap into careers where they'll be miserable, or turn up their noses at jobs they would find rewarding. They need careers advice.
The challenge of this project is to develop an artificial intelligence application to provide that careers advice. It should ask questions and collect test scores of various kinds, and apply some combination of rule based reasoning and numerical calculations to figure out what careers would be good and bad choices.
Personality and aptitude tests and principles for choosing careers come in all shapes and sizes. One approach to this problem is to build a modular system that applies different methods or rule sets to giving advice - which may conflict! Another approach is to consider how different information might be synthesised to provide a good overall rating for a possible career.
Variant
An application for such an approach might be to compute rankings of the suitability of a pool of applicants for a range of jobs, for instance the staffing requests received by a temping agency, or placements offered to/by a university sandwich year placement unit.
Back to