What is a model?
There are two good reasons for having models of information retrieval. The first is that models guide research and provide the means for academic discussion. The second reason is that models can serve as a blueprint to implement an actual retrieval system.
Mathematical models are used in many scientific areas with the objective to understand and reason about some behaviour or phenomenon in the real world. One might for instance think of a model of our solar system that predicts the position of the planets on a particular date, or one might think of a model of the world climate that predicts the temperature given the atmospheric emissions of greenhouse gases. A model of information retrieval predicts and explains what a user will find relevant given the user query. The correctness of the model’s predictions can be tested in a controlled experiment. In order to do predic- tions and reach a better understanding of information retrieval, models should be firmly grounded in intuitions, metaphors and some branch of mathematics. Intuitions are important because they help to get a model accepted as reason- able by the research community. Metaphors are important because they help to explain the implications of a model to a bigger audience. For instance, by comparing the earth’s atmosphere with a greenhouse, non-experts will under- stand the implications of certain models of the atmosphere. Mathematics are essential to formalise a model, to ensure consistency, and to make sure that it can be implemented in a real system. As such, a model of information retrieval serves as a blueprint which is used to implement an actual information retrieval system.