Analysing terms, pairs, triplets and full queries used in intranet searching

Paper in proceeding, 2007

Web search engines have become often-used tools for many ordinary people today and a growing number of researcher are therefore studying how these lay-persons interact with such tools. Studies of public web search engine usage have often produced term frequency lists to illustrate the information needs the users. This study differs on several aspects from previous work. Firstly, we have analysed the logs of an intranet search engine, since studies of corporate internal search behaviour are in short supply. Secondly, we have not just used search terms but also full queries and show that single terms give a skewed understanding. Thirdly, we have analysed data from three different years - 2000, 2002 and 2004 - to be able to detect shifts and trends in information seeking behaviour.