An approach to cluster users on Last.fm
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
.vscode
01_data_fetcher
02_data_preparation
03_data_processing
04_data_ploting
__pycache__
data_processed
visualizations
.gitignore
DM-2017-report-stoecklmair.pdf
README.md
data.z01
data.z02
data.z03
data.z04
data.z05
data.z06
data.z07
data.z08
data.z09
data.z10
data.z11
data.zip
presentation.key
thoughts.md

README.md

  1. First fetch artists, tags and tracks

    1. ./01_data_fetcher/fetch_top_artists.py
    2. ./01_data_fetcher/fetch_top_tracks.py
    3. ./01_data_fetcher/fetch_top_tags.py
    4. ./01_data_fetcher/fetch_users_top_artists.py
    5. ./01_data_fetcher/fetch_users_recent_tracks.py
    6. ./01_data_fetcher/fetch_users_top_tracks.py
  2. Generate Files

    1. ./02_data_preparation/UsersGen.py
    2. ./02_data_preparation/AristsGen.py (depending on ./data/users.txt)
    3. ./02_data_preparation/TracksGen.py (depending on ./data/artists.txt)
    4. ./02_data_preparation/UserRecentTracksGen.py (depending on ./data/tracks.txt)
    5. ./02_data_preparation/UserTopTracksGen.py (depending on ./data/tracks.txt)
    6. ./02_data_preparation/UserTopArtistGen.py (depending on ./data/tracks.txt)
  3. Fetch Data Again

    1. ./01_data_fetcher/fetch_artists_meta.py (depending on ./data/artists.txt)
  4. Generate Files Again

    1. ./02_data_preparation/TagsGen.py (depending on ./fetched_data/artists_meta/*.json)
    2. ./02_data_preparation/FillArtists.py (depending on ./fetched_data/artists_meta/*.json)
  5. Generate processed data to analyze

    1. ./03_data_processing/top_tags-user_top_artists.py - Get top tags and count how often a user heard that artist (user top artists)

Notes

  • Every page of user_recent_tracks starts with the value nowplaying: true
  • Tags to lower case and remove [ -] to combine e.g. Hip hop, hip hop and hip-hop

Presentation

  • Ansatz
  • Was gemacht
  • Ergebnisse
  • Was gelernt
  • Deutsch/English
  • Report in English (4-8 Seiten)
    • Guideline: Nachimplementierung sollte nach dem Lesen möglich sein (nachvollziehbar)