Just few days ago the popular ReadWriteWeb published a list of the 2009 Top Ten Semantic Web products as they did one year ago with the 2008 Top Ten.
This two milestones are a good opportunity to make something similar to a balance. Or just to do a quick overview on what’s changed in the “Web of Data”, only one year later.
The 2008 Top Ten foreseen the following applications, listed in the same ReadWriteWeb order and enriched with some personal opinions.
Yahoo Search Monkey
It’s great. Search Monkey represents the first kind of next-generation search engines due its capability to be fully customized by third party developers. Recently, a breaking news woke up the “sem webbers” of the whole planet: Yahoo started to show structured data exposed with RDFa in the search results page. That news bounced all over the Web and those interested in SEO started to appreciate Semantic Web technologies for their business. But, unfortunately, at the moment I’m writing, RDFa is not showed anymore on search results due to an layout update that broke this functionality. Even if there are rumors on a imminent fixing of this, the main problem is the robustness and the reliability of that kind of services: investors need to be properly guaranteed on the effectiveness of their investments.
Probably, this neat application has became really popular when it has been acquired by Microsoft. It allows to make simple natural language queries like “film where Kevin Spacey acted” and, a first glance, the results seems really much better than other traditional search engines. Honestly I don’t really know what are the technologies they are using to do this magic. But, it would be nice to compare their results with an hypothetical service that translates such human text queries in a set of SPARQL queries over DBpedia. Anyone interested in do that? I’ll be more than happy to be engaged in a project like that.
With a large and massive branding operation these guys built the image of this service as it be the only one fitting everyone’s need when dealing with semantic enrichment of unstructured free-texts. Even this is partly true (why don’t mentioning the Apache UIMA Open Calais annotator?), there are a lot of other interesting services that are, for certain aspects, more intriguing than the Reuters one. Don’t believe me? Let’s give a try to AlchemyAPI.
I have to admit my ignorance here. I never heard about it, but it looks very very interesting. Certainly this service that offers, mainly, some sort of semantic advertisement is more than promising. I’ll keep an eye on it.
Down at the moment I’m writing.😦
Many friends of mine are using it and this could be enough to give it popularity. Again, I don’t know if they are using some of the W3C Semantic Web technologies to models their data. RDF or not, this is a neat example of semantic web application with a good potential: is this enough to you?
Another case of personal ignorance. This magic is, mainly, a restaurant review site. BooRah uses semantic analysis and natural language processing to aggregate reviews from food blogs. Because of this, BooRah can recognize praise and criticism in these reviews and then rates restaurants accordingly to them. One criticism? The underlying data are perhaps not so much rich. Sounds impossible to me that searching for “Pizza in Italy” returns nothing.
Blue Organizer (or GetGlue?)
It’s not a secret that I consider Glue one of the most innovative and intriguing stuff on the Web. And when it appeared on the ReadWriteWeb 10 Top Semantic Web applications was far away from what is now. Just one year later, GetGlue (Blue Organizer seems to be the former name) appears as a growing and live community of people that realized how is important to wave the Web with the aim of a tool that act as a content cross-recommender. Moreover GetGlue provides a neat set of Web APIs that I’m widely using within the NoTube project.
A clear idea, a powerful branding and a well designed set of services accessible with Web APIs make Zemanta one of the most successful product on the stage. Do I have to say anything more? If you like Zemanta I suggest you to keep an eye also on Loomp, a nice stuff presented at the European Semantic Technology Conference 2009.
Mainly, a semantic search engine over a huge database containing more than 400,000 hotels in the US. Where’s the semantic there? Uptake.com crawls and semantically extracts the information implicitly hidden in those records. A good example of how innovative technologies could be applied to well-know application domains as the hotels searching one.
On year later…
Indubitably, 2009 has been ruled by the Linked Data Initiative, as I love to call it. Officially Linked Data is about “using the Web to connect related data that wasn’t previously linked, or using the Web to lower the barriers to linking data currently linked using other methods” and, if we look to its growing rate, could be simple to bet on it success.
Here is the the 2009 top-ten where I omitted GetGlue, Zemanta and OpenCalais since they already appeared also in the 2008 edition:
Google Search Options and Rich Snippets
When this new feature of Google has been announced the whole Semantic Web community realized that something very powerful started to move along. Google Rich Snippet makes use of the RDFa contained in the HTML Web pages to power rich snippets feature.
It’s a very very nice feeds aggregator built upon Google Reader, Twitter and FriendFeed. It’s easy to use, nice and really useful (well, at least it seems so to me) but, unfortunately, I cannot see where is the Semantic aspects here.
BBC Semantic Music Project
Built upon Musicbrainz.org (one of the most representative Linked Data cloud) it’s a very remarkable initiative. Personally, I’m using it within the NoTube project to disambiguate Last.fm bands. Concretely, given a certain Last.fm band identifier, I make a query to the BBC /music that returns me a URI. With this URI I ask the sameas.org service to give me other URIs referring to the same band. In this way I can associate to every Last.fm bands a set of Linked Data URIs where obtain a full flavor of coherent data about them.
It’s an open, semantically marked up shared database powered by Metaweb.com a great company based in San Francisco. Its popularity is growing fast, as ReadWriteWeb already noticed. Somehow similar to Wikipedia, Freebase provides all the mechanisms necessary to syndicate its data in a machine-readable form. Mainly, with RDF. Moreover, other Linked Data clouds started to add owl:sameAs links to Freebase: do I have to add something else?
DBpedia is the nucleus of the Web of Data. The only thing I’d like to add is: it deserves to be on the ReadWriteWeb 2009 top-ten more than the others.
It’s a remarkable US government initiative to “increase public access to high value, machine readable datasets generated by the Executive Branch of the Federal Government.”. It’s a start and I dream to see something like this even here in Italy.
So what’s up in the end?
It’s my opinion that the 2009 has been the year of Linked Data. New clouds born every month, new links between the already existent ones are established and a new breed of developers are being aware of the potential and the threats of Linked Data consuming applications. It seems that the Web of Data is finally taking shape even if something strange is still in the air. First of all, if we give a closer look to the ReadWriteWeb 2009 Top Ten I have to underline that 3 products on 10 already were also in the 2008 chart. Maybe the popular blog liked to stress on the progresses that these products made but it sound a bit strange to me that they forgot nice products such as the FreeMix, Alchemy API, Sindice, OpenLink Virtuoso and the BestBuy.com usage of GoodRelations ontology. Secondly, 3 products listed in the 2009 chart are public-funded initiatives that, even if is reasonable due to the nature of the products, it leave me with the impression that private investors are not in the loop yet.
What I expect from the 2010, then?
A large and massive rush to using RDFa for SEO porpoises, a sustained grow of Linked Data clouds and, I really hope, the rise of a new application paradigm grounded to the consumption of such interlinked data.