\input zb-basic
\input zb-ioport
\iteman{io-port 06008656}
\itemau{Kiefer, Peter}
\itemti{Mobile intention recognition. Foreword by Christoph Schlieder.}
\itemso{Berlin: Springer (ISBN 978-1-4614-1853-5/hbk; 978-1-4614-1854-2/ebook). xvi, 166~p. EUR~79.95/net; SFR~106.50; \sterling~72.00; \$~99.00 (2012).}
\itemab
Foreword: Understanding what the user of a mobile information system plans to do next -- recognizing his or her intentions -- is of crucial importance for the design of location-based services (LBS). Solutions to the mobile intention recognition problem are especially needed in application scenarios where the opportunities to interact with the mobile device via a keyboard or a touchscreen are limited. This is the case for the use of smart-phones in many outdoor activities such as riding a bike, skiing, or running. Even hikers prefer not to stop to interact with the device. Ideally, the LBS would analyze the spatial behavior of the user, identify the user's intentions to act, select the currently active intention, and finally provide the information services that assist the user in achieving the intended goal.  In his thesis, Peter Kiefer provides the first comprehensive treatment of the field. This includes not only a thorough review of the state of the art but also a formal characterization of the mobile intention recognition problem as opposed to the general intention recognition problem. Although the approach is formal, it is deeply rooted in the practical experience that the author of this book gained while designing LBS, especially tourist guides and location-based games.  The challenge of intention recognition consists in finding an adequate approach for representing the background knowledge about the structure of the spatial environment (e.g. the partonomic structure of a city) and the spatiotemporally constrained set of possible actions of the user. It is far from clear how to best combine this background knowledge with the behavioral data. Some obvious research issues are: Which spatial and temporal restrictions provide the relevant constraints for recognizing intentions in LBS? What is an adequate representational formalism for these constraints? Which computational mechanisms permit to solve the recognition problem? The author's thesis gives convincing answers to all these questions.  Here are some of the results which I think are particularly interesting: (1) a generic layered architecture for mobile services that perform intention recognition, (2) an analysis of behavioral pattern which shows that rule-based representation formalisms of different expressive power are needed to handle the problem, (3) two grammar formalisms, spatially constrained context-free grammars (SCCFGs) and spatially constrained tree-adjoining grammars (SCTAGs) which come with appropriate parsing algorithms that solve the intention recognition problem, (4) last, but not least, the implementation of the generic architecture as a framework for evaluating different approaches to intention recognition.  Peter Kiefer has written this thesis with the idea of presenting his research results in a larger context. In turns out that he succeeded writing a very readable book that is of value to those interested in LBS in general and intention recognition in particular. Because of the comprehensive and knowledgeable survey of the state of the art it may also serve as an introductory text to the field that is accessible not just to computer scientists but also to researchers from geographic information science.
\itemrv{~}
\itemcc{}
\itemut{location-based services; mobile information systems; mobile intention recognition; intelligent mobile assistance systems}
\itemli{doi:10.1007/978-1-4614-1854-2}
\end