Hackers donate 90% of profit to charity 2019/06/13

NGI Zero awarded two EC research and innovation actions 2018/12/01

EC publishes study on Next Generation Internet 2025 2018/10/05

Bob Goudriaan successor of Marc Gauw 2017/10/12

NLnet Labs' Jaap Akkerhuis inducted in Internet Hall of Fame 2017/09/19




Search engines are the default way of finding information on the internet. Although there is a host of search engines for users to choose - from library catalogs to cooking portals - there is currently only a small number of dominant search engines that practically decide who finds what on the internet. This situation has the following disadvantages: 1) by designing their algorithms these dominant search engines influence our world view, 2) the huge amounts of user data they record, creates sever risks of data leaks and misuse, finally 3) search engines can misuse their market power to gain advantages in other lines of business (e.g. the mobile phone market).

Federated web search is a technology where users connect to a so-called broker which forwards their search request to suitable search engines and combines the results. Using federated search lessens the risks of few dominant search engines: it shows a blend of search results created by different algorithms, it prevents the search engine to record data of individual users, and its search results are usually more divers. Still, for federated web search to become widely used, it faces the following challenges: 1) while exploiting user behavior is known to improve search effectiveness, brokers exploiting this data also risk leaks and misuse, 2) as brokers typically serve many users, they are not able to include search engines for personal content, such as email, social media or cloud storage because the public broker cannot know the user’s credentials to access these services, finally 3) brokers consider for every user the same base set of search engines, while considering a more focused set of engines could improve search results, given the diversity of users.

To improve upon these challenges, while avoiding the disadvantages of dominant search engines, this project will investigate a radical change to the federated search architecture: users run a broker on their own computer using a browser plugin. In this architecture the broker can safely analyze the user's behavior to improve search results as the data is accumulated on a per-user basis on disconnected computers. Furthermore, the search requests forwarded to search engines use the user's credentials and thus can access search engines for personal data, such as email etc. Finally, starting from sensible defaults, each user can configure its broker with his or her individual needs.

Logo NLnet: abstract logo of four people seen from above Logo NGI Zero: letterlogo shaped like a tag

This project was funded through the NGI0 Discovery Fund, a fund established by NLnet with financial support from the European Commission's Next Generation Internet programme, under the aegis of DG Communications Networks, Content and Technology under grant agreement No 825322. Applications are still open, you can apply today.


Send in your ideas.
Deadline October 1st, 2019.


Last update: 2019/05/15