Adriaan de Groot, the Dutch psychologist and chess Master, argued that “perception and memory are more important differentiators of chess expertise than the ability to look ahead in selecting a chess move” (Groot 1978). A component of expertise in chess has been attributed to the expert having knowledge of ‘chunks’ and this knowledge gives the expert the ability to focus quickly on “good moves with only moderate look-ahead search” (Gobet and Simon 1998). The effects of chunking in chess are widely reported in the literature, however papers reporting the nature of chunks are largely based on inference from psychological experimentation. This thesis reports original work resulting from extensive data mining of a large number of chessboard configurations to explore the nature of chunks within the game of chess and the associated moves played by expert chess players. The research was informed by work in the psychology of chess and explored with software engineering techniques, employing large datasets consisting of transcripts from expert players games. The thesis reports results from an analysis of chunks throughout the game of chess, explores the properties of meaningful chunks and reports effects of the application of chunk knowledge to move searching.