A longitudinal study of information needs and search behaviors in science and technology: A query analysis

This study aims at a longitudinal understanding of the user–system interactions from the context of science and technology at a query level.

The authors quantitatively analyzed log data sets culled from more than 24,820,416 queries submitted by users of a national scientific and technical information system, collected in 2008-2011.

In the fields of science and technology, the user search behaviors and patterns have remained stable. User queries are short and simple. In all, 80 per cent of the queries are made up of one-three terms. The length of query on a scholarly information system in the fields of science and technology is different from that of Web search. The former is longer than the latter. Search topics have shifted fast. “FUEL BATTERY”, “NANO”, “OLED”, “CAR”, “ROBOT” and “SMARTPHONE” were high-ranked queries from 2008 to 2011. It was found that the time to determine whether the users will stay on the site took about 10 seconds on average from the time of visit. If the users viewed the results of a list generated by the search query and took any action, such as detailed view, export or full-text download, most of them stayed more than 10 minutes on average.

Longitudinal user research using a query analysis helps to understand the information needs and behavioral patterns of users on information systems related to a specific field and those based on the Web. It also brings insights into the past, present and future events of a field. In other words, it plays a role as a mirror that reflects the flow of time. In the long run, it will be an historic asset. In the future, user studies using a query analysis need to be carried out from various (e.g. social, cultural or other academic disciplines) long-term perspectives on a continuous basis.