I was searching through the E-prints in Library and Information Science (E-LIS) web site for search related articles over the past year, and came across some articles that weren’t highly cited, but provide some interesting perspectives on Google Scholar, and on research involving user behavior on the web and how to understand it better.
Examining the Claims of Google Scholar as a Serious Information Source
by Bruce White (09 November 2006)
A thoughtful and detailed analysis of the strengths and weaknesses of Google Scholar from a member of the library community. Bruce White’s conclusion points to the use of Google Scholar as an essential informational resource but not a one-stop-shop for searching academic journal literature.
Google Scholar Citations and Google Web/URL Citations: A Multi-Discipline Exploratory Analysis
by Kayvan Kousha and Mike Thelwall (05 June 2006)
A second paper that takes a good look at Google Scholar (and Google Web search), and compares them to a more traditional scientific literature citation resource.
The study focused upon locating scientific literature citations across a wide range of disciplines, and determined that Google Scholar and Google Web search were useful and effective sources for finding information that might not have otherwise been seen – especially in the social sciences area, and in fields where it is common to publish conference papers online, such as in computer science.
Web User Tasks and Search Behavior
A Goal-based Classification of Web Information Tasks
by Melanie Kellar, Carolyn Watters, and Michael Shepherd (08 January 2007)
People used a specialized web browser that recorded their activities on the web, including how they interacted with the browser’s navigational features. They also recorded information about their activities in an electronic diary. The idea was to get a better sense of how people performed different tasks on the web.
What Can Searching Behavior Tell Us About the Difficulty of Information Tasks? A Study of Web Navigation
by Jacek Gwizdka and Ian Spence (05 December 2006)
Finding information on the web is a task oriented endeavor, and this study attempts to compare objective measures with users’ perceived notions of how easy or difficult it is to perform specific search-related tasks.
Investigating the Performance of Automatic New Topic Identification Across Multiple Datasets
by H. Cenk Ã–zmutlu, Fatih Cavdur, Amanda Spink, and Seda Ã–zmutlu (16 December 2006)
Early studies looking at users’ search behaviors focused upon one query at a time. Viewing search sessions in log files provide more information about how people are using search engines, but present a few challenges, such as when people switch search tasks during a single search session.
It isn’t always easy to determine when someone switches from one search task to another during a single session. Can this be done with the help of automated processes and neural networks? What factors in a search engine database might make a difference in pinpointing those changes.
This study looks at the processes involved while using two different search engines – FAST and Excite, to see if a neural network can be trained to effectively understand topic changes in search sessions while using data from both.