Getting to grips with squishy boolean
Lizzie Sparrow was intrigued by the term after hearing it used at Internet Librarian International, and decided to find out more
Squishy boolean is a term I came across in Marydee Ojala’s presentation at the Internet Librarian International conference at Olympia last month. I loved the sound of the term (cuddly boolean, how fun!) but wasn’t entirely sure what it meant. I received the same sentiment from others on twitter so decided to check it out.
It seems squishy boolean was a term coined in 2005 by Mary Ellen Bates¹ and refers to a range of search algorithms less rigid than boolean, such as 'preferably includes', relevancy ranking, using approximate matches to your search terms and personalised search results.
What really hit me reading this article was that Bates was talking about squishy boolean giving the searcher more control. The idea of the user choosing if they wanted to use this squishy boolean, and then choosing how they wanted to use it. Whereas, what we now have in most search systems is squishy boolean imposed upon us whether we want it or not, and often there is no way of finding out what algorithms have been used.
To me, this is a real shame. Sometimes, squishy boolean is great. For example, I can type the word 'boots' into Google, and right at the top of the results is a map of Cambridge (my hometown) with pins for each of the Boots stores and their opening time beneath. Google successfully translates my one word into the not exactly simple question: 'When and where can I get to a Boots shop?', understanding that I don’t want to travel a long way and the shop needs to be open.
On the other hand, there are times when I don’t want to do a ‘normal’ search. Perhaps I’ve got a more complex search to do. I’ve found various techniques to get around Google’s assumption that it knows best: putting search terms in quotation marks; using – when I want to search for x NOT y; searching for specific file types or in certain domains.
But these techniques aren’t easy to find and those available are different for different search interfaces. And when I don’t want personalised results I’ll use a different search engine (if you want some ideas of what to use check out Phil Bradley’s website).
I was discussing how Google personalises search results with a couple of non-library colleagues last week and it scares them. They feel uncomfortable with not understanding google’s magic black box and feel they should use something else that doesn’t ‘steal’ their data, but don’t really want to leave the convenience and familiarity of Google. They were keen to learn about different options open to them, and wanted to be able to make an informed choice about when and if to use different search engines. Whatever type of library we work in, this is something librarians can offer users advice on. It’s all part of digital information literacy.
At Internet Librarian International Marydee Ojala predicted that web search will get better, but not for librarians. It may get better at working out what we ask the majority of the time, but I’m not sure the web search experience will get better for anyone. Privacy concerns and the ever decreasing transparency of search engine workings do not create a happy searching public. We (librarians) need to help.
1Bates, M. E. (2005) online spotlight: Squishy Boolean. Online 29.2: 64
For those anyone with CILIP membership you can access this article via LISA. From the CILIP homepage go to Membership –> Benefits –> Monthly magazine, journals and ebulletins –> Online journals and select the ejournal collection/database of your choice. Bates’ full article is available via Library and Information Science Abstracts (LISA).
This article was initially published on Lizzie Sparrow's blog, here. Lizzie is the librarian at BirdLife International, a global nature conservation partnership, and interim librarian at the Cambridge Conservation Initiative, setting up a library and information service for their new conservation campus.
