valiantone
commented
5 months ago Good proposal, well documented @cnymz. I think the only open questions would be on DS/DL side. You seem to be creating a lot of the metrics using community, but a huge factor in project success is in being able to acquire event/concert data very fast. Without this applying DS will be difficult and even BE would end up waiting a long time for their data.
- Have you confirmed the existence of concert datasets in addition to the music categorical datasets?
- Do you plan to only use spotify datasets or even live info using spotify API? Both are acceptable options.
On UX and WD sides, task is straight-forward and we do not see any hurdles. If you just respond to the DS concerns, we can start discussing team formation and move the project to Accepted.
cnymz
What is your project's name?
Community-powered concert tracker/recommender
What's the problem you are trying to solve?
Berlin is a big city, full of events, especially musical ones. Although there are apps/platforms that list concerts by "well-known" artists, it's very difficult to find events by lesser-known artists, amateurs...
The problem is all the more difficult when your musical tastes are particular and correspond to a music scene that doesn't have a wide audience. Generally speaking, to access this kind of information, you already need to be connected to a community organized around a specific genre.
The problem obviously works the other way round. Less-known artists find it difficult to connect with a wider audience or even with their potential audience.
Many of these events take place in bars, restaurants, associations, schools or as part of one-off events. They can be concert, mini-concert, jam-session, open-mic etc., mostly free events or events which do not have ticket sales but a small on-site entrance fee. This is, by the way, one of the reasons why these events are not listed, or are poorly listed, by existing platforms.
In the end, information is difficult to access, it requires too much effort. Although motivated, we find ourselves missing out on a lot of concerts.
What's your idea for a solution to the given "problem"?
A platform which lists all the concerts and musical events which are not, or are poorly listed by existing platforms. A platform with the goal of connecting the right concert, at the right time, at the right place to the right person. A platform with two main features:
Which tracks do you think could be involved?
DL/DS, WD, UX
How do you imagine Deep Learning and/or Data Science could contribute to this project?
We have huge Spotify datasets available in which we have access to different characteristics to define an artist or a song. These can be criteria such as genre or subgenre, release date, country etc. but even more interesting are certain attributes such as: acousticness, danceability, duration_ms, energy, instrumentalness, key, liveness, loudness, mode, speechiness, tempo, time_signature, valence.
By applying ML models on these datasets, it is theoritically possible to find classification criteria allowing us to develop a relevant classification/search/recommendation system on our platform to connect an artist/event to an audience/user. The use of ML represents one of the two main components of the project.
Being in the DS track myself, I have so far only been able to think within the framework of this type of skills. But obviously, there would surely be more features to develop involving DL. The project is completely open to proposals in this direction.
How do you imagine User Experience Design could contribute to this project?
The project proposed here aims to be able to be used by as many people as possible, we will need UX designers in the project in order to make the platform use simple and instinctive.
How do you imagine Web Development could contribute to this project?
By creating the plateform :) The project would be more relevant in mobile app format. I don't know if that would be possible.
What are the most important challenges & bottlenecks of your project?
Quickly agree as a team on the main outlines of the MVP Be able to deploy the proposed machine learning solution. If possible develop the project as a mobile app. To have a simple and fun user interface
What are the most important success factors of your project?
For the success of the platform: organic communication, offer the platform to all musician communities in Berlin. A free communication tool for events can only be welcomed if it is easy to use. The musicians' entourage and the musicians themselves will be enough to form the community of active early adopters.
Share your resources here:
https://www.kaggle.com/datasets/jfreyberg/spotify-chart-data https://www.kaggle.com/datasets/rodolfofigueroa/spotify-12m-songs https://www.kaggle.com/datasets/nikitricky/every-noise-at-once https://everynoise.com/ https://spotipy.readthedocs.io/en/2.19.0/#