The workshop in Paris has been a great opportunity to take a quick and refreshing dive into one of the most exciting EU-funded projects on Semantic Web technologies.
We at InSideOut10 (running this project in partnership with Interact SpA, one of the companies I co-founded in the mid-nineties to help Public Institutions as well as Private companies developing their own footprint on the Web) are participating as early-adopters and we’re working on a WordPress Plug-In code-named WordLift to help bloggers around the world “spoon-feeding” HTML Microdata to Search Engines (following the May 2011 announcement of Schema.org – you can find more about this topic on last Wernher’s post).
What do we love…about IKS?
Following the launch of Google latest revamped UI called What do you love? I decided to present here a list of aspects of IKS I found intriguing and worthwhile at the same time:
- IKS really is a Treasure Chest (and this is not mine but David’s first definition): a collection of open source technologies to semantically enhance text by recognising named entities, and by categorising them as places, organisations or people (IKS is more then that but for the purpose of this article let’s keep this definition); the key to the open the chest is in the hands of the braves who decide to approach it (researcher, technology provider, public institution, private companies or independent developers) and how they leverage to use it;
- The tools provided are worth the time you need to understand (let alone installing and configuring) an incubating and ever-changing open source framework like IKS;
- An EU-funded project that has sucessfully entered the Apache Incubator (and it’s therefore preparing its entrance into the Apache Software Foundation) is a step ahead of most of the on-going research in Europe in this field in terms of providing to a large industry sector (CMS providers in this case) an enterprise-grade freely available open source software as deliverable of the research program. More over the interoperation within others ASF projects is already providing some benefits at the technical level (OpenNLP, Clerezza and Solr/Lucene are the projects that seemed to me closer to IKS scope).
The gold-rush: is Semantic Web already a business opportunity?
As an entrepreneur my contribution to this question is crucial: my interest on the Semantic Web technologies has to meet my clients’ business requirements and their path to innovation.
IKS is a collaboration between Academia and Industry and yet during the workshop (as I suspected) the interests of Academia seemed to outweight the needs of business to deliver compelling solutions. The opportunities that Semantic Web is unleashing were, for my liking, pitched a bit too optimistically (again with the eye of an Internet Entrepreneur); I’ll come back to this point in a future blog – with recommendations on how to bridge the gap between research and business scenarios.
As far as my experience goes in this field I found 3 successful approaches when it comes to create an actual business value with Semantic Web technologies:
- 1. Content analysis and content strategy: especially in large corporations a key issue is delivering the right message to the right target; a lot of the investments in a company’s external relations department go to media monitoring services, social media analysis and competitive studies. These reports are typically lacking knowledge on the gap between what a company wants to communicate and what has been said by others on the same topic. Using text-mining techniques and leveraging the potential of content enrichment made possible with semantic web technologies can provide a much clearer picture of the overall scenario – in real time. Moreover this analysis of the “semantic gap” can provide valuable insight to improve the content marketing strategy, such as create synergies within otherwise separated information silos.
- 2. Transparency and data marketing: disclosing company data can be beneficial in various way: the airline industry demonstrates how opening up data did help their overall business (though it took a while). They came to realize that having their flight data on more sites and in more searches increased their business. Here is article on this topic http://thinkquarterly.co.uk/01-data/open-for-business/ – it does require more effort of course to convince the private sector to release their dataset but as some one recently said “data is the new oil”.
- 3. Leveraging search engine marketing investments: content enrichment and semantic mark-up (like RDFa or HTML microdata) do increase the content findability hence providing a cost effective way to attract more search traffic; considering existing SEM investments by large corporation this is an effective way to support semantic technologies.
This is our motivation to open the Treasure Chest of IKS. The experience has proved to be successful (though we’re still at the very early stage, note that I’m not talking here about an overview of the market or a study but about real business case scenarios).
The road ahead: where I expect to see IKS 2 years from now
IKS started in January 2009 and has a duration of 48 months hence it will be over by end of 2012, embracing Apache Software Foundation in November 2010 as mentioned above was a great move to address the overall project sustainability (there is an interesting article on this topic written by Seth Grimes who was at the Paris Workshop) and attract various open source communities.
This might not be enough, a fruitful collaboration with Industry is a must for the project to survive and to face the upcoming challenges; the support of big players (could Adobe become a major sponsor of IKS once the EU funding ends?) as well more traction in the rising web 3.0 eco-system could bring new life to IKS beyond 2012 – and this is what I wish we will see.
At the technological level I expect to see the following:
- Strong focus on non-english languages (this is a strategic goal – a very interesting demo was shown in Paris by Olivier Grisel on document categorization and during the hackaton David with Oliver’s advice successfully implemented the categorization using the Italian DBPedia on Italian content for the first time – this was exciting!)
- Support to fully untap the potentials of automatic reasoning (big move, today this is still way too experimental)
- Entity relationship extraction (currently not yet available in the Treasure chest other then directly through OpenNLP).
Once again a big thanks to the whole IKS community!