Dear developers and data enthusiasts, we are excited to announce our very first DBpedia challenge!
Wikipedia presents a way to easily access knowledge in whichever language we need, and its infoboxes provide useful information at one glance. Unfortunately, at times there is some discrepancy in the facts presented in the infoboxes of the different wikis. Creating a tool to allow the user to detect these discrepancies would significantly improve the quality of the information provided by Wikipedia. And considering that there are other sources of knowledge on the web - such as Wikidata, DBpedia, and other databases - their implementation would allow for Wikipedia to be an even more powerful knowledge hub.
What is the challenge?
If we presented you with a database containing information from different sources, how would you go about to create such a tool to bring consistency to the infoboxes of the individual Wikipedias? We want to see your realization of such a mechanism to help create even more accurate Wikipedias.
Prizes & Perks
You will gain unlimited access to the DBpedia pre-fusion database, and we will announce the top three projects in a DBpedia blog post, providing a chance to be seen by the DBpedia community. The participant with the best realization will be presented a paid 6 month internship at the DBpedia Association.
Which data sets will be available?
- DBpedia Pre-fusion database
- DBpedia Databus dump if a more complex set of data is needed (in this case we could also provide a server to accommodate extensive calculations/data processing)
- Maybe give a sync target to limit the possible outcome?
- Access to the database will be provided via email
- Send to who?
- Dbpedia student members (via Slack channel)
- GSoC students (via Sandra)
- Post on twitter & FB & August newsletter