Computing Reviews

Interactions with search systems
White R., Cambridge University Press,New York, NY,2016.Type:Book
Date Reviewed: 09/06/16

Web search has become so ingrained in modern life that is hard to imagine what we would do if it suddenly disappeared, say by some unforeseen catastrophe. While a large fraction of our searches are simple lookup transactions that can be answered factually (for example, the location of the nearest restaurant or the birth data of a famous person), a good number of them are fairly complex and require us to interact with the search engine, often over multiple sessions spanning a long period (for example, finding the “ideal” location for a vacation somewhere hot or the “best” hiking boots for people with wide feet). In these cases, several lookup transactions are performed, interspersed with possibly unrelated cognitive tasks by the users, before the information need is satisfied.

Interactions with search systems by Ryen W. White compiles an impressive body of research spanning several decades of work looking into how humans interact with search systems and how such interactions affect both the efficacy of the search (as in the ability of the system serving highly relevant information) and also the process through which the user performs the search itself. Particularly for information needs that require long sessions spanning multiple individual queries, such interactions can have a significant impact on the outcome. Some examples of this point discussed in the book include searching for medical information, particularly symptoms of serious diseases or medical conditions, and the way in which the snippets of the answers are crafted by the search engine (for example, whether or not to show such symptoms).

The research described in this book is complex, interdisciplinary in nature, and very difficult to carry out in a scientific way. It is hard to model search interactions beyond simple lookup transactions, especially the cognitive processes performed by users while interacting with the system. Furthermore, there are several technical challenges in instrumenting web browsers with tracking capabilities that can reliably offer interaction data, especially for long-running search sessions. Obviously, there are several serious privacy concerns associated with the collection of such data, even if the purpose of providing better search experiences is well justified. Finally, the challenges in evaluating such research cannot be overestimated. Not to mention that the advent of novel interaction modes introduced by industry often drastically change the landscape, requiring even more research in the area.

White’s book covers quite a lot of ground and many disparate topics in its approximately 400 pages excluding references and notes. While it will hardly become anyone’s favorite bedtime reading, it will certainly be of great use to researchers and graduate students interested in search systems and how humans interact with them. More than that, the book offers quite a lot of inspiration for researchers working on related areas that are not strictly search in the sense of information retrieval. For example, I was left thinking that many of the ideas could be applied to the study of how students “search” for answers to problem sets. All in all, this is great addition to the libraries of researchers and students working in the area.

Reviewer:  Denilson Barbosa Review #: CR144730 (1612-0873)

Reproduction in whole or in part without permission is prohibited.   Copyright 2024 ComputingReviews.com™
Terms of Use
| Privacy Policy