The methodology of information retrieval experiment

Information retrieval systems have been the subject of experimental testing for some twenty years now. Like any field in this position, a fair amount of know-how has accumulated about the proper conduct of such investigations. The object of this book is to distil this know-how; the object of this chapter is to set the scene. Thus I will be introducing the basic ideas, sketching in some of the main problem areas, and generally preparing the reader for the more specific or concrete chapters that follow. Van Rijsbergen takes up the question of the evaluation of retrieval effectiveness in Chapter 3; in Chapter 4, Belkin considers information retrieval in a wider context; and in Chapter 5, Tague gets down to the detail of conducting experiments.