Last updated on 2015-05-28
About the Book
Music services (Spotify, iTunes radio, Pandora, etc.) need to stand out; yet, this is getting harder since most catalogs overlap, besides a few exclusives. And, as online music becomes be a commodity, they need to find incentives for users to use them versus competitors.
One way to do so is to invest more time -- both on product and R&D -- on the technology front, especially on personalisation and discovery, in order to be ahead of the pack and own the space. This is an obvious strategy, and a win-win-win for all parties involved:
- For consumers, delighted when they discover new artists they will love -- based on their past listening habits or the ones of their friends --; and satisfied as they figure out that streaming services really understand what they like;
- For artists, escaping the long-tail and hence generating more streams, and a little revenue, but most importantly: having the opportunity to convert casual listeners into super-fans who follow them on tour, buy merch and exclusive records, and more;
- For streaming services, keeping existing users active with more listening hours and growing their audiences; consequently gathering more data and analytics (plays, thumbs-up, social interactions, etc.), which can be re-invested into product features.
This e-book, #MusicTech, is an evolving summary of various hacks and experiments on data-science, recommender systems - with a focus on online streaming, and more globally analytics and music discovery.
Using tools such as Semantic Web technologies, Big Data infrastructures, Machine Learning, Collaborative Filtering and more, you will learn different techniques to make sense of music streaming data, featuring use-cases which aim to be fun and entertaining, and cover well-known platforms, such as Twitter, Spotify, or YouTube.
Knowledge graphs, discovery, and personalization on the Web
- From knowledge graphs to content recommendations
- From knowledge graphs to taste graphs, and to personalization
Sex and drugs and Rock’n’roll: Analytics of the Rolling Stone’s 500 greatest songs of all time
- Come together
- Baby love
- I wanna be anarchy
- Hotel California
- Good vibrations
Back in Black: A focus on the loudness and tempo of the Rolling Stone top 500 songs
- Black leather, knee-hole pants, can’t play no high school dance
- But you don’t really care for music, do you?
- Please could you stop the noise, I’m trying to get some rest
- I’m waiting for that final moment you say the words that I can’t say
The role of mood and tempo for music discovery?
- You Can’t, You Won’t And You Don’t Stop
- Fear of the Dark
- Mood, tempo, and personalization
Driving music recommendations through Wikipedia edits
- Wikipedia edits as a data source
- The dataset
- Querying for similarity using pages edits
- Towards long tail discovery
- Going further
Gathering insights from 500,000 Deezer playlists
- Setting up the stage
- Content recommendations
- Top artists and tracks, popularity, and more
- Trends, influencers and targeted recommendations
How YouTube music is shared on Twitter
- Popular videos: Super fans or spammers?
- Entities: Better than tags
- Twitter, personalization and music
- Knowledge graphs, discovery, and personalization on the Web
An open stack for music personalization on the Web
- The growth of dynamic and structured data
- Personalizing content from multiple sources
- Privacy matters
Context: The future of music streaming and personalization?
- How context and wearables will enhance music discovery
- Building the future of context-based music personalization
- An open stack for music personalization on the Web
The Leanpub 45-day 100% Happiness Guarantee
Within 45 days of purchase you can get a 100% refund on any Leanpub purchase, in two clicks.
See full terms...