by lasagna_coder on 1/7/22, 2:15 PM with 25 comments
The methods I'm aware of are:
- Recommendation engines (e.g. Youtube, most social media feeds)
- Organic / constructed sequences of data (e.g. a playlist, or sorted lists using some measurement like price)
- Non-text search (e.g. search via image, sound, or geolocation)
- Stochastic (e.g. guessing a domain name, or entering some random text into a search engine and skipping 100 first suggestions)
- Automatic link traversal (e.g. a web crawler algorithm)
- Private sharing (e.g. sending a link in an email)
- Public sharing (e.g. web portal, or blogroll)
- Hybrid (e.g. algorithmic search engine results, news aggregators)
But what am I missing?
by gostsamo on 1/7/22, 2:36 PM
by manx on 1/7/22, 7:07 PM
by sargstuff on 1/8/22, 6:24 PM
** recommender system conferences:
recsys -> https://dl.acm.org/doi/proceedings/10.1145/3240323
umap -> https://dl.acm.org/doi/proceedings/10.1145/3450613
sigweb -> https://www.sigweb.org/conferences/acm-sigweb-conferences
** "Recommender Systems Handbook" with source code : https://link.springer.com/book/10.1007/978-0-387-85820-3
by freediver on 1/8/22, 12:33 AM
It automates the process of discovery based on previous interest.