So what have we learned from this?
Developing applications as part of a research project
Vines has been quite a different activity compared to other parts of the IKM program. Most other activities were desk studies and workshops. For Vines computer applications have been developed. They have been discussed at workshops and have played part in the overall linked data activity. But most of the effort went into overcoming the usual development problems (such as incompatible versions of custom modules). With hindsight it would have been better to have more modes design goals and invest more effort in working with communities.
We found that it was very hard to find appropriate newsfeeds with content from the South. The newsfeed from Codesria journals was created as part of this effort. (Codesria has since then migrated to another system) Further discussions at the workshop that we held in Dakar confirm the conclusion that the first hurdle to make the South more visible on the Internet is getting the content on the web in the first place. Organisations like INASP and IMARK work on capacity building in this respect. Especially the experience with Euforicvines shows that online content from the South does not turn up prominently enough in search engines. So capacity building should also include some basic notions of search engine optimization.
The experience with regard to terminologies is limited because only a limited number of information sources have been used. If we compare Euforicvines and Codesriavines it is striking to see that not just the words are different (which they are obviously because they come from different list) but also the underlying concepts seem to be quite different. Maybe the most striking is the fact that in the material from Codesria the financial crisis was already an important issue, while it was not used bur the documentalists from the North who added tags to Southern material for Euforicvines
The number of real-life applications of semantic technologies, although there are some (like the BBC site or the Vivo system that Cornell University has developed and is now being rolled out in other universities in the USA) What has been developed for IKM (both Young Lives and Vines 2) are proof-of-concept applications to introduce the issues in the development community. IKM semantic provides a relatively simple method to map terminologies for use in the semantic web.