IMSLP folks:
I'm interested in helping performers discover compositions that
1) appeal to their individual taste, and
2) are within their technical range.
These could be obscure pieces by old composers,
or new compositions by contemporary composers.
Platforms like Spotify and YouTube have personalized mechanisms
that let listeners discover new music;
we need something analogous for classical music performers.
To this end, I've developed a web site, Classical Music Index (CMI):
http://34.83.223.220/cmi_web/
(sorry, no domain name or SSL yet).
Like IMSLP, CMI lets you search for compositions based on metadata.
CMI is seeded with a snapshot of IMSLP data from 2023,
converted from Wiki to a database format;
CMI doesn't store scores or recordings.
CMI has features for music discovery:
- You can rate and review items (compositions, composers, performances, etc.)
You can rate the technical difficulty of a composition as well as its musical quality.
- When you view an item you can see its ratings and reviews,
the users who wrote them, what other items they reviewed, etc.
You see a 'people who liked this item also liked...' list.
- After CMI gets a critical mass of ratings,
it will provide recommendations using 'collaborative filtering':
mining the rating data using statistical and AI techniques
to predict a particular user's rating of a particular item,
and (in the case of compositions) its difficulty for that user.
CMI has some other novel features.
It lets you add and edit items via an error-resistant web interface.
Instead of typing things, you choose from drop-down lists.
Fields are automatically error-checked.
Also, CMI can store information about concerts.
CMI is my idea of what IMSLP should become.
If you have time, I'd appreciate your feedback on CMI.
Create an account, search for things,
rate them, write reviews, creates new items if you want.
Let me know if you have questions or problems,
and tell me what you think of CMI:
the ideas and goals behind it, and the web site UI/UX.
Reply to this post, or email me (davea@berkeley.edu)
or create an issue in the CMI Github repository.
Thanks -
David P. Anderson
University of California, Berkeley
Ideas for enabling music discovery
-
- Posts: 1
- Joined: Mon Nov 28, 2022 8:23 pm
- notabot: 42
- notabot2: Human