search

Tribler / tribler

Privacy enhanced BitTorrent client with P2P content discovery
https://www.tribler.org
GNU General Public License v3.0
4.74k stars 445 forks source link

My updates: recommended based on your downloads and searches #1846

Open synctext opened 8 years ago

synctext commented 8 years ago

Enhance the frontscreen of Tribler with content items in the form of thumbnails. These items are recommended based on popularity in general and fit with your profile. Suggested GUI text: "My updates: recommended based on your downloads and searches"

Key inspiration (include in thesis.tex): scholar_my_updates_screenshot__personal_recommendation

Related: #1549, https://github.com/Tribler/tribler/pull/981

working code #634 screenshot

Technical documentation: #372 Docs

synctext commented 8 years ago

Youtube also has a nice thumbnail-based navigation system. Key example for the Tribler User Experience. Youtube features:

Each recommendation or profile entry can be easily deleted. wjifjbq

btw.. It is very aggressive to recommend new videos, one accidental "mylittlepony" click can ruin your recommendations. This problem was widely published already in 2002, If TiVo Thinks You Are Gay, Here's How to Set It Straight

whirm commented 8 years ago

You could always have a button to clear your preferences.

synctext commented 5 years ago

Delft University of Technology scientists published in RecSys '18 Proceedings of the 12th ACM Conference on Recommender Systems: "Effects of Personal Characteristics on Music Recommender Systems with Different Levels of Controllability" recommender_explaining Very relevant work for this open issue.

Our system generates a play-list style listening experience based on three types of seeds: artists, tracks, and genres. We use the active user’s top artists, tracks, and genres as input seeds. It is worth noting that the top artists and tracks are calculated by affinity, which is a measure of expected user preference for a particular track or artist based on her/his listening history. The number of songs recommended through the use of a particular seed depends on the weight of the seed’s type, and the priority of the used seed among the seeds of the same type.

As more bands publish using Creative Commons or put their work into Tribler and obtain 100% of all tokens, this could be the way to discover music. Very suitable for a master thesis.

synctext commented 4 years ago

related work: economics of content recommendation, Media and Artificial Intelligence, Matthew Gentzkow, http://conference.nber.org/conf_papers/f114672.slides.pdf

Sources of inefficiency

  1. Consumers can’t find what they want o Imperfect search, matching, recommendations
  2. What consumers want isn’t what’s good for society o Fake news, bias, Kardashians, violence
  3. Gov’t, firms, etc. have other ideas o Censorship, capture, foreign manipulation, persuasive ads