The Tony Kent Strix Award is an annual award for outstanding contributions to the field of information retrieval. It is presented in memory of Dr Tony Kent who made a major contribution to the development of information science and information services in the UK and internationally.
The Award itself has been granted annually since 1998 but this year the saw the first of what will be an annual lecture series. Sponsored by Google, the lecture series will feature the previous year’s winner of the Award.
Last year’s winner, Susan Dumais of Microsoft Research, gave the inaugural lecture at London’s Geological Society on 6th November 2015. She chose to explore how we can understand and improve web searching by collecting and analysing large scale behavioural data.
User behaviour logs help us better understand the 'real world' behaviour of searchers. Data can be collected using lab studies (controlled tasks or think aloud protocols for example); panel studies or log studies.
Early web search log research found that the top 250 search queries accounted for 10% of all searches and that this was followed by an incredibly diverse 'long tail' of queries which might only appear once. These insights show that search engines simply must get the top 250 searches 'right'.
The why behind the what
Logs can help us understand search behaviour, but they do not offer insight into user intent. For example, search abandonment may be a bad sign - or a good one (e.g. the answer being sought appears in the inline/snippet text and no further clicking was required).
The objective of search is to help people accomplish tasks and to 'delight' them with results. Just as search engines are learning from users, so users are 'learning' from search engines. The next field of research may well be to explore how users are faring with spoken-word searches and results.
For more information on the Tony Kent Strix Award, see the UKeiG Awards and Busaries page.