London Enterprise Search Meetup – Databases vs. Search and Taxonomies

Back to London for the next Enterprise Search Meetup, this time featuring Stefan Olafsson of TwigKit and Jeremy Bentley of Smartlogic.

Stefan started off with a brief look at relational databases and search engines, and whether the latter can ever supersede the former. He talked about how modern search technologies such as Apache Solr share many of the same features as the new generation of NoSQL databases, but how in practise one often seems to end up with a combination of search engine and relational database – an experience we share, although we have a small number of customers who have entirely moved away from databases in favour of a search engine.

Jeremy’s talk was an in-depth look at Smartlogic’s products, which include taxonomy creation and management tools, and are designed to complement search engines such as Solr or the GSA. Some interesting points here including the assertion that ‘we trust our content to systems that know nothing about our content’ – i.e. word processors, content storage and management systems – and that we rely on users to add consistent metadata. Smartlogic’s products promise to automate this metadata creation and he had some interesting examples such as the NHS Choices website.

Some interesting discussions followed on the value of taxonomies. Our view is that open taxonomy resources such as Freebase are better than those developed and kept private within organisations, as this can prevent duplication and promote cooperation and the sharing of information. Also, taxonomies often seem to be introduced as a way to fix a broken search experience – maybe fixing the search should be a higher priority.

Thanks to Tyler Tate for organising the event – the tenth in this series of Meetups, and now a regular and much anticipated event in the calendar.

Share this post

Facebook

Google+

Twitter

Linkedin

Reddit

email