Earlier this month I attended the third Enterprise Search Europe conference, this time not to speak but to run workshops, panels, tracks and social events. On Tuesday a colleague and I gave a workshop on Getting the Best from Open Source Search which I hope was useful to attendees: one thing I did take away is how the level of experience with open source and indeed search technology itself can vary widely: some attendees had already experimented widely with Apache Lucene/Solr and some simply wanted to expand their knowledge of the associated risks & benefits of this approach.
The first day of the conference started with Ed Dale of Ernst & Young talking about implementing enterprise search for a truly global organisation. E&Y’s search is over a surprisingly small number of documents (only 2 million or so) but they are lucky enough to have a relatively large and experienced team running their search as an ongoing operation – no ‘fire and forget’ here (an approach often taken and seldom successfully). We moved on to hear from Kristian Norling on the second year of Findwise’s Enterprise Search Survey (some interesting numbers with the full results available soon) and then a fascinating and amusing talk from Joe Lamantia on the Language of Discovery, backed up by a second talk from Tyler Tate – it seems Discovery might a better term for what we call Search, at least from a usability perspective. The morning ended with Steven Arnold’s provocative take on how the performance of search technology hasn’t improved measurably in many decades due to processing limitations and how the rise of Big Data is only going to compound the problem.
The afternoon began with a panel session on the future of open source search – my personal thanks to Daniel Lee of Artirix, Eric Pugh of Open Source Connections and René Kriegler for leading a lively discussion on the seemingly inexorable rise of open source search and what may happen next. There were some interesting points raised on how significant investment in open source search may change the picture. We continued in the open source theme with talks on open source solutions for the City of Antibes and Shopping24, before a drinks reception and then moving to the pub across the road for the combined London and Cambridge Search Meetup. Our theme was ‘The Nightmare before Search’ – some great (and unbloggable!) war stories on crazy search implementations was followed by networking late into the night.
The next day continued with a session on search implementation from speakers including Dan Foster of Legal & General, a track on Big Data during which we heard from Eric Pugh on building a very large scale system using open source software – sadly I had to drop out at this point for meetings and only returned for the closing plenary sessions. I particularly enjoyed Kara Pernice‘s insights on how to build usable intranet search and Valentin Richter’s session on migrating to a new search technology (a topic on many minds especially for those using FAST ESP which goes out of mainstream support in a couple of months). Lynda Moulton did her best to sum up what we had learnt over the last few days – a very hard job when the event covered so many aspects of search & discovery.
Many thanks to Information Today and chair Martin White as ever for organising the event – although it was an intense few days it was great to catch up with everyone and to talk search. We’re looking forward to next year – did I hear a rumour that the Europe in the title might be more emphasized next time? We shall see!