Applied Informetrics for Information Retrieval Research (review)

Both information retrieval (IR) and informetrics apply quantitative measures to investigate underlying regularities of scholarly discourse. This book presents a significant attempt to intersect the two areas, with the assumption that the conceptual linkage and common approaches will enable unprecedented insights into phenomena in information studies.

The author argues that informetrics shares a conceptual foundation with the study of IR. While informetrics focuses on mapping properties of scholarly communication networks, typical objects of research are surrogates of scholarly publications, such as journal articles and books stored in an IR system. IR research is aimed at developing algorithms for effective retrieval and implementing systems that provide access to information resources. Researchers in IR and informetrics are both interested in using quantifiable indicators to describe, explore, and predict information transfer patterns in various stages of the information production process (or IPP, as per the author). Despite much overlap and research proximity, "to date informetrics and information retrieval have been largely treated as separate areas of investigation, with occasional research forays into specific informetric aspects of IR." (p. 3)

In advocating the application of informetrics to IR research, the author presents a comprehensive review of concepts and theories in both areas of study. The description of each aspect of research is concise yet understandable. As a result, even though the book contains mathematical equations and notations, the content is not intimidating for advanced undergraduate and graduate readers. Several interesting ideas include the IR process model as illustrated on page 13, and the comparison of database management systems (DBMS) and IR systems. The IR process model describes the interaction components in IR and is a clear integration of several previous IR models. The comparison of DBMS and IR is long overdue and should be helpful for Library and Information Science (LIS) students since frequently LIS courses cover the two systems without differentiation.

The book covers a rich set of mathematical models, ranging from classic models to the most recent measures. On the IR side, both traditional models (such as Boolean, Vector Space, and Probabilistic) and evaluation measures (such as recall, precision, effectiveness, and response time) are presented. Information visualization, distributed and cross-lingual IR, and Z39.50 standardization are also discussed. On the informetrics side, the author describes principles of Lotka's, Bradford's, and Zipf's Laws. Calculations of impact factor, co-citation clustering, co-word analysis, and measures of literature growth, obsolescence, and S&T indicators are also presented. The exhaustive coverage makes the book a useful reference source for both teaching and research.

In later chapters, the author indicates informetrics findings can be and in reality already have been embedded in IR research. The author foresees potential applications in simulation of IR systems, IR process design and implementation, and IR system evaluation. An informetric approach will gear IR research toward an orientation to use and users, as well as foster profound, progressive, and scientifically solid research fronts that carry great implications to IR system development.

The key issue in assessing the true value of the book is what it has to offer compared to several established IR books including Baeze-Yates and Ribeiro-Neto's Modern Information Retrieval (ACM Press, 1999), Korfhage's Information Storage and Retrieval (Wiley Computer Publishing, 1997), and Meadow, Boyce, and Kraft's Text Information RetrievalSystems (Academic Press, 2000). Obviously, informetrics has not been thoroughly handled in those works. Additionally, some recent concepts and findings are only covered in this book. On the other hand, some descriptions of relevant topics seem sketchy in this book, and discussion sometimes stops prematurely and refers to other books and articles for further readings. For instance, the summary on models of browsing and information seeking is somewhat superficial. The author fails to mention social network analysis software frequently used by bibliometric researchers. Two important growth measures—citations outside category (COC) and Brillouin's diversity index—are not included. The informetrics chapter includes no coverage of discourse analysis or text linguistics. The research technique sections do...