@MASTERSTHESIS{pgi2020010,
	author = "A. Russell",
	supervisor = "Y. Moshfeghi",
	title = "A Grenetic Alogrithm for Query Optimization",
	school = "Department of Computer and Information Sciences, University of Strathclyde",
	year = "2019",
	abstract = "The growing quantities of unstructured textual information online can represent an
untapped goldmine of data, but one of the barriers to exploiting it is {\"\i}{\neg}nding where
these resources are. Information retrieval systems and search engine technologies have
developed to tackle this need. However, there are elements of them that can be further
developped to suit our needs. In particular, the problem of user queries not being
optimal for the search they are trying to perform persists. If we could optimize the
query being provided to a retrieval system this could go a long way to improving the
quality of information users are receiving from search engines.
In this dissertation the applicability of a genetic algorithm to query optimization within
the context of information retrieval is explored. First and foremost the goal is to inves
tigate whether this is an e{\"\i}{\neg}€ective way of altering a search query to improve retrieval.
However, the research is done within the context of trying to develop technologies that
would allow easier research and data collection via internet search engines. This spe
ci{\"\i}{\neg}c context requires a unique experimental design framework where the information
retrieval system, and the document collection it has indexed, provide limited informa
tion to the query expansion techniques being applied.
This report shows that a genetic algorithm approach to query optimization proves
more e{\"\i}{\neg}€ective than other query expansion techniques for retrieval, and introduces a
framework for performing such query optimization applicable to any search engine.",
}