Speak of the devil… build your own OpenCalais like supermachine

In line with my recent blog mentioning OpenCalais, the topic extraction tool, DBpedia, one of the awesome linked open data projects I’ve been using a bunch for Alive.cn, just released their own topic extraction tool, DBpedia Spotlight. If you are okay with downloading 9GB of Lucene indices and setting up their scripts, you can have your own self-hosted topic extraction tool. They basically open sourced something that is worth a lot of money in a previously relatively closed space.

What is topic extraction? Check this demo out and enter any block of text — say, a recent news article. The benefit of using DBpedia’s solution (besides it being free) is that it automatically ties topics back to their DBpedia topics which already have a huge storehouse of Wikipedia-derived linked open data.