Book Review: Information Representation and Retrieval in the Digital Age
There are two distinct aspects covered by this book: information representation and information retrieval. It is refreshing to see a book seek to address these together, as it is important to recognise how interlinked they are. As the author herself indicates in the preface, ' - two chapters of this book are devoted exclusively to information representation, a step that must be taken before information becomes retrievable'. These two chapters certainly bring a fresh perspective to considerations of information retrieval, coming as they do at the start of the book and influencing later discussion. The author herself suggests that the book should be read sequentially and although this is generally true the chapters can be referenced in their own right if the reader has some background knowledge.
Having said that, those with some knowledge will find more comprehensive and potentially more useful sources elsewhere. The author does also suggest that this book will be most useful to those who are newly interested in learning about the field. In this she has mainly succeeded, as many basic aspects and areas are described; it is clear that Professor Chu is coming from a teacher's background, wishing to introduce new students to issues in the field of information representation and retrieval (IRR). That is the field of IRR that is focused on research and theory. Readers of this book will get a good grasp of underlying IRR concepts and research in the area; they will not get any great overview of the practical application of these concepts or trends in the exploitation and practice of them. This is a shame, as where specific examples are given to explain a concept the meaning shines through far more clearly. But there are too few such examples.
It is also a shame that having set up the importance of information representation alongside retrieval, that this area is then treated rather meagrely. The two chapters certainly lay out the ground, but there was much detail not covered. For example, the section on metadata would leave readers with a rather skewed and underdone understanding of this essential area. This seems a missed opportunity as information retrieval is covered in far greater depth; links are made subsequently to the relationship between representation and retrieval, but these are on the surface and are not followed up. There is a suspicion that Professor Chu's area of expertise is information retrieval, and representation has been included as a side issue.
The other aspect included in the title is that of IRR in the digital age. This is certainly addressed and the discussion of applicability of information retrieval (note, not information representation) concepts to Internet search engines is useful. There is again, however, not as much integrated discussion of how the digital age, or more accurately the Internet in this book, has affected IRR, which again seems a missed opportunity.
These missed opportunities should not, though, undermine what is welcome in this book. The chapters on language in IRR, retrieval techniques (for the most part), retrieval models, and retrieval of information unique in content or format (again, in parts) are very useful summaries, if at times somewhat dependent on particular key sources. The author is at her best when discussing the theoretical basis of information retrieval. The evaluation of IRR systems is also covered well, albeit that what the chapter really covers is the evaluation of information retrieval systems, not representation. When she moves outside these clearly more comfortable areas the text is not so helpful. The chapter on the user dimension of IRR manages to avoid mention of usability at all, albeit that the issues covered are certainly relevant. The section on the use of artificial intelligence appears an add-on at the end, not really reflecting the level of work in this area.
In summary, this book sets out its aims and meets them on the whole. But it is frustrating that a book that appeared to want to address the connections between information representation and retrieval has only scratched the surface. This may, of course, reflect the lack of research and work on this connection, in which case more is needed, but along with other areas of this book there are gaps that do not appear to have been filled. This may push new students onto other texts in these areas, and I hope it does. So long as they don't feel let down by where they started.