Search User Interfaces [Englisch] [Gebundene Ausgabe]

This book is a survey of recent work in search, but with an unusual focus on the importance of interface design on searcher's perceptions of the quality and usefulness of search results.

It is an excellent supplement to core search texts such as Introduction to Information Retrieval (Manning et al., 2008) which focus on the backend technology behind search. Search User Interfaces focuses more on the how we search, what we expect to see when we search, different interfaces that have been tried in the past, and which of those people found useful.

More specifically, the author starts by looking at why Google and other search engines have such a spartan design, explaining that searchers find it most helpful when distractions are minimized and it is quick and easy to iterate on searches. She goes on to lay a foundation by looking at the many models for how people search, including an exploration of the information foraging model where searchers partially satisfy some goals while rapidly developing new goals as they are exposed to new information. She offers hints on techniques that have worked well (e.g. immediately showing search results, keywords-in-context in the snippets, diversity of results on ambiguous queries, biasing results based on query term order and proximity, the importance of seemingly minor design tweaks, just to name a few). She dismisses more complicated interfaces such as boolean queries, thumbnails of result pages, clustering, pseudo-relevance feedback, explicit personalization, and visualizations of query refinements and search results, saying that they showed poor results in the past. She holds out hope for faceted search, universal search, and implicit personalization.

I enjoyed the way this book usefully points at techniques which have shown promise while dismissing others as consistently confusing to users. It is a guide to what works and what does not in search, warning of paths that likely lead into the weeds and pointing to better opportunities.

Professor Marti Hearst has written the comprehensive review of all current research on search engine user interfaces. But this book is more than just a survey of the literature: it explains all aspects of successful search use interface design (and usability and user experience). Many of us have had ideas that seem nice but simply do not work in practice. Reading this book will save everyone involved with search engine design a huge amount of time and trouble.

The writing in this book is extremely clear and direct: it's an ideal textbook for anyone interested in search engines in general as well as interface issues. I wish it had been around when I was first (painfully) learning many of these lessons. Highly recommended.

I confess that before reading this book I had not heard Hearst's name, but a quick search of the literature in this area clearly shows her to be central to this field (and my lack of knowledge of the field in not having heard her name). Dr. Marti Hearst is a Professor in the School of Information at the University of California, Berkeley and was previously a member of the research staff at the highly respected Xerox PARC. Reading between the lines of the book I hope you will see a person dedicated to her work and undoubtedly world class in her field. Even if this were not the first book in this field, it would be hard to find something more definitive in the area. Hearst masterfully weaves together academic work across time and from around the world, and is clearly on the ball in terms of current developments. She sets her own work humbly, but in its rightful place in the history of search user interfaces.
The structure of the book itself is a testament to her ideas and principles of design. From the outset we find a clearly written contents page followed by a book overview which neatly encapsulates the books content, setting each chapter in context. Hearst also provides advice on her intended audiences. It's easy to navigate the book and it can be read cover to cover, or dipped into for knowledge and insight in a particular area.
Drawing widely from the literature and using clear examples Hearst begins the book by introducing some general ideas and practices surrounding interface design. Chapter two is a natural follow on from user-centred design as it tackles evaluation of search interfaces. The information seeking process is the subject of the third chapter where Hearst looks at models of the process. The next three chapters cover the information seeking cycle with a look at query specification, the presentation of search results and query reformulation. Hearst then looks at interfaces that support the information seeking process in Chapter seven. The important topic of integrating navigation with search is highlighted in Chapter eight, with the emergence of personalization in search the topic for Chapter nine. Two chapters on aspects of Information visualization then follow and the book is capped off with a discussion of emerging trends in search interfaces.
I fall firmly into the category of her second intended audience, the practitioner, so what value did I derive from the book? Firstly, this is not a dry academic book, it's alive! The work really drew me in, and at one point I found myself wondering what it would be like to work with Hearst. Reading the book I experienced an information parallel to the Californian Gold Rush, dashing out of the information mine with another nugget of knowledge to apply. I'm at a loss to find anything to `criticize' with this book. The nature of the material makes it a dense read, but that's the complexity of the field rather than Hearst's lucid style. One thing which would augment the book would be some associated teaching resources which would help to make the best of this work as a practitioner. In this fast moving field, I'm looking forward to seeing a second edition.