Towards a personalized Internet: a case for a full decentralization

Abstract

The Web has become a user-centric platform where users post, share, annotate, comment and forward content be it text, videos, pictures, URLs, etc. This social dimension creates tremendous new opportunities for information exchange over the Internet, as exemplified by the surprising and exponential growth of social networks and collaborative platforms. Yet, niche content is sometimes difficult to retrieve using traditional search engines because they target the mass rather than the individual. Likewise, relieving users from useless notification is tricky in a world where there is so much information and so little of interest for each and every one of us. We argue that ultra-specific content could be retrieved and disseminated should search and notification be personalized to fit this new setting. We also argue that users’ interests should be implicitly captured by the system rather than relying on explicit classifications simply because the world is by nature unstructured, dynamic and users do not want to be hampered in their actions by a tight and static framework. In this paper, we review some existing personalization approaches, most of which are centralized. We then advocate the need for fully decentralized systems because personalization raises two main issues. Firstly, personalization requires information to be stored and maintained at a user granularity which can significantly hurt the scalability of a centralized solution. Secondly, at a time when the ‘big brother is watching you’ attitude is prominent, users may be more and more reluctant to give away their personal data to the few large companies that can afford such personalization. We start by showing how to achieve personalization in decentralized systems and conclude with the research agenda ahead.

Peer-to-Peer System Design Trade-Offs: A Framework Exploring the Balance between Blockchain and IPFS

“…The second wave of decentralized solutions was achieved through fully distributed technology; i.e., P2P networks without classical servers but instead using ordinary computers (different from classical cluster/grid parallel computing). There have been multiple attempts to offer P2P web services [24,25], such as Freenet for censorship-resistant communication [26], although broad adoption has mostly been limited to the field of file sharing; e.g., eDonkey, BitTorrent [27].…”

Open Peer-to-Peer Systems over Blockchain and IPFS

“…Full decentralization would be certainly useful, especially for certain applications [30]. However, it was not until recently that some unresolved technical challenges [33,45] have become more evident, which have been the driving forces to innovations such as Blockchain [39] and IPFS [3].…”

Web science: a new frontier

“…Kermarrec's [7] paper describes new kinds of distributed architecture for the Web. It is driven by the observation that personalization is a dominant theme on the Web.…”