Elasticsearch for the win

I stole that heading from the inspiring Andrew Rowe, senior developer at the Western Australian Museum. Andrew and I worked together for about 5.5 years, and before I left we started a disusion about decoupling the presentation layer and search from the backend.

Good news! Andrew has just informed me of an exciting development back at the WAM - placing an AngularJS and Elasticsearch UI on top of CollectiveAccess. With the best possible results.

I've been investigating and working with CollectiveAccess for a while now (see my previous blog on running CollectiveAccess in Vagrant if you want to have a play with it) but I have not always enjoyed the speed of the search when it scales. In fact, I have regarded it as its worst feature. It has a nasty tendancy to not give feedback when you start faceting search (as in the processing and drilldown occurs without a visual feedback queue to indicate your request is being processed) - which on a big query, say faceting several hundred thousand results - can take a really, really long time.

Andrew has solved this problem with the help of ElasticUI. Now results return immediately, and as you type queries, results automatically appear. Bang! Google-like predictive completion within a museum collection. Now that is open access to a collection, using open tools. I look forward to sharing more developments as it they come out (and hopefully at some point in the future, access to the public as well).