Music Recommendation - From Peel To Peer

Video thumbnail (Frame 0) Video thumbnail (Frame 1683) Video thumbnail (Frame 2300) Video thumbnail (Frame 11094) Video thumbnail (Frame 12903) Video thumbnail (Frame 14749) Video thumbnail (Frame 18456) Video thumbnail (Frame 19232) Video thumbnail (Frame 20063) Video thumbnail (Frame 20841) Video thumbnail (Frame 21414) Video thumbnail (Frame 22472) Video thumbnail (Frame 24884) Video thumbnail (Frame 26946) Video thumbnail (Frame 32943) Video thumbnail (Frame 33995) Video thumbnail (Frame 34728) Video thumbnail (Frame 36989) Video thumbnail (Frame 38072) Video thumbnail (Frame 41092) Video thumbnail (Frame 41986) Video thumbnail (Frame 42624) Video thumbnail (Frame 43441)
Video in TIB AV-Portal: Music Recommendation - From Peel To Peer

Formal Metadata

Music Recommendation - From Peel To Peer
Curatorial Models
Title of Series
Number of Parts
CC Attribution - ShareAlike 3.0 Germany:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor and the work or content is shared also in adapted form only under the conditions of this license.
Release Date

Content Metadata

Subject Area
Music recommendation is an essential part of the music business. Artists like Trent Reznor have started to get involved in music disovery systems and talk of "Intelligent Curation". It might be time to envision a more emotional, i.e. relationship based approach to recommendations. Are we entering the age of "intelligent curation"? Just like the distribution of music has undergone fundamental changes in the last 15 years, so has the culture of music recommendation. The development of recommendation engines is an essential aspect of online retailers and music platforms, while social media has all but replaced traditional media as the dominant recommendation channel, resulting in a new ecology and economy of taste making.
Peer-to-peer Peer-to-peer Causality Term (mathematics) Directed set Endliche Modelltheorie Musical ensemble Form (programming) Window
Multiplication sign Source code Set (mathematics) Open set Mereology Bookmark (World Wide Web) Mathematical morphology Web 2.0 Inference Programmer (hardware) Mathematics Different (Kate Ryan album) Computer configuration Hypermedia Videoconferencing Endliche Modelltheorie Logic gate Descriptive statistics Physical system Block (periodic table) Software developer Moment (mathematics) Demoscene Degree (graph theory) Message passing Sparse matrix Search algorithm Order (biology) Right angle Text editor Figurate number Quicksort Row (database) Web page Trail Pay television Computer file Link (knot theory) Wage labour Virtual machine Letterpress printing Field (computer science) Twitter Power (physics) Latent heat Thermodynamisches System Internetworking Software testing Computing platform Metropolitan area network Form (programming) Matching (graph theory) Forcing (mathematics) Mathematical analysis Expert system Volume (thermodynamics) Library catalog Exploit (computer security) Peer-to-peer Newsletter Word Personal digital assistant Blog Einbettung <Mathematik> Musical ensemble Abstraction
Standard deviation Scaling (geometry) Multiplication sign Range (statistics) Source code Independence (probability theory) Bit Mereology Degree (graph theory) Goodness of fit Limit set Cuboid Speech synthesis Text editor Musical ensemble Freeware Row (database) Physical system
Ocean current Type theory Term (mathematics) Multiplication sign Moment (mathematics) Order (biology) Musical ensemble Quicksort Mereology Freezing
Divisor Computer file Multiplication sign Login Broadcasting (networking) Bit rate Strategy game Term (mathematics) Different (Kate Ryan album) Internetworking Videoconferencing Series (mathematics) Endliche Modelltheorie Computing platform Distribution (mathematics) Key (cryptography) Software developer Mathematical analysis Electronic mailing list Bit Arithmetic mean Visualization (computer graphics) Software Musical ensemble Local ring
Distribution (mathematics) Service (economics) Digitizing Energy level Independence (probability theory) Whiteboard Musical ensemble Family Associative property Formal language
Presentation of a group Theory of relativity Moment (mathematics) Mathematical analysis Shared memory Content (media) Independence (probability theory) Online help Word Different (Kate Ryan album) Hypermedia Musical ensemble Endliche Modelltheorie
Slide rule Algorithm Service (economics) Software developer Independence (probability theory) Mereology Number Bit rate Term (mathematics) Endliche Modelltheorie Quicksort Musical ensemble Physical system
Mixture model Divisor Observational study Sound effect Self-organization Musical ensemble Quicksort Data conversion
Web page Beat (acoustics) Finitismus Service (economics) Divisor Direction (geometry) Numbering scheme Set (mathematics) Akkumulator <Informatik> Twitter Formal language Product (business) Endliche Modelltheorie Theory of relativity Moment (mathematics) Electronic mailing list Mathematical analysis Bit Data mining Arithmetic mean Integrated development environment Personal digital assistant Speech synthesis Musical ensemble Quicksort Figurate number
Turing test Medical imaging Service (economics) Touchscreen Musical ensemble
Source code 1 (number) Sound effect Bit Lattice (order) Drop (liquid) Instance (computer science) Streaming media Degree (graph theory) Latent heat Strategy game Musical ensemble Quicksort Endliche Modelltheorie Row (database)
Discrete group Context awareness Block (periodic table) View (database) Multiplication sign Source code Virtual machine Bit Online help Streaming media System call Different (Kate Ryan album) Personal digital assistant Blog Musical ensemble Endliche Modelltheorie Extension (kinesiology)
Stiff equation Broadcasting (networking) Different (Kate Ryan album) Weight Interactive television Electronic mailing list Software testing Musical ensemble Endliche Modelltheorie Computer programming
Broadcasting (networking) Bit rate Workstation <Musikinstrument> Sampling (statistics)
but it the huge and
it and then hello and welcome
stage form the next 2 sessions
are on the topic of music recommendation by both take a different perspectives the first one um as were uh organized by constant term the stem and ends accustomed will be joined by Orchid gutless later on and it it focuses on the curation aspect of music recommendation that session that follows after this focuses more on that Tecnologica and research as by Don of recommendation and both sessions have been recomended stood public our by the Berlin-based snatch work as sociation all together now and but now we start with Erik was Stam who has down with the topic but she cause from peer-to-peer um for the past few years and am has
thought a lot about how music recommendation where it's now differently in the digital age than it did before because down is uh the writer uh radio host and there's lots of different things and welcome thank you and there and this work in the microphone checked yes it works only at my name is Chris and how messages occurred
and on the session analysis starts with a little introduction by myself before I happily introduce my partner for today could greatly and um later so will be discussed for more recent developments and platforms in the field of situated recommendation but I would like to start with some opening remarks in a more general thing and also I'd like to thank all together now and especially on the and for making it possible for us to be here the traditional from peer-to-peer is of something that started in 2009 or together now and I like to call it a look at the ecology of music particularly the the change in music recommending at the beginning of this I thought that was the man versus machine kind of situation with the opponents being music experts vs. search algorithms meanwhile a lot of that has changed like my opinion about this as well and mostly that has to do with the success of social media and the various options for streaming music and embedding these things 1 radical changes that the so-called human music recommendations in the web now mostly occurs links and made it mostly that lead to the instant consumption of the respective music so there is no more the need for abstract descriptions of something that can be heard straight away and recommendation even if it's not taking place inside 1 of the music platforms themselves has grown much nearer to the actual experience music in most cases so the referring used in social media forces hyping up some new Track Act video would be very similar to the referring employed in nineties hyping when it was done in print so not a big change there but the message is mostly spreading much quicker and the music he heard instantly another change is the sheer volume of music that is being released are accessible by now if you include that catalogs and the respective availability of also not so recent or even a and sounds on the streaming platforms the amount of problems tracks Complete Works lost sessions mixtapes bootlegs and so forth can be quite intimidating but obviously we're not alone out there is held in form of all sorts of recommendations and for now we want to look at the so called she rated models once you have the infamous culture of the gatekeepers powerful figures in a system of restricted test that watched over the stars we sources of so called tell I'm talking of radio hosts magazine editors TV show programmers and their respective counterparts in the recording business like in almond mostly these were men promoters another set of in order to be accessible by a greater public music had to be let through gates that achievement financial backing and promotion in the fairly closed system of the old new school music business 1 could argue about the nurturing benefits the system offered for the lucky few and not all of it was there but it is known the system is now defined in many ways and the gates have mostly opened or are there many many new arrests for unguarded gates and doors and loopholes there's no Moscow city of music on the sparsity of time and attention and started to have money somewhat and we talk more also influences now maybe then of course in the good influences are not a specific breed of people or have a specific profession In many ways there were the so-called immaterial labor something between amateur enthusiasm semi-professional says exploitation and what is called motion work done on behalf of Internet media bronze or other employers but you can decide which inferences you letting for you at all these influences can be a good friend who always has provided you with new music depending on your birthday by mixtapes make CDs or some later form of sharing but inferences can also musicians whose music you like music journalism is you know from traditional media channels bloggers of collective blocks deejays flaps labels communities are some mistake referential figure in a file way music scene that we noticed some other people's tweets it everyone can theoretically or rather quite easily build themselves a system of social media page lights and friends of newsletter subscriptions or subscriptions of video on Twitter tumbler SoundCloud of interest or whatever channels and start getting more recommendations and you can manage not even mention hours as feeds from music dexterity such a system of self-selected sources if done smartly will almost always guarantee an astonishing degree of matches with your own taste and you will always find new music right the moment you assess the respective channel and once you start sharing your favorites from all this will obviously become an influenza yourself later on the morphologies of concept of the district users may be more fancy word for that as well and that sounds pretty easy and also the easy part in the URI but it's
not so easy part and that is that this whole takes quite a bit of an effort 1st of all you have to be interested enough to do all this and you have to have time to focus your attention on the new music that floats in boxes and timelines for dashboards whatever you look at probably on the scale of the system you use will reflect the degree to which your cultural practices and this is refined basically and that's not too much of a surprise more effort would get if you more distinctive gain so if you're not willing to spend much time and effort in finding good sources for recommendation would have to make do with a limited set of channels just subscribe to and probably have a bit of a lesser amount of discoveries of exciting new sounds but if you're ready to engage with the wider range of influences and influences began find itself better supplied with new music now many music editors of the past received a pile of records or cds each day delivered to the best for we used by these editors and I'm speaking from experience were forced to occupy themselves all sorts of unwanted but personally delivered music you can now choose freely from the
whole week he for his music and only listen to what you preselect and you can even start to develop a certain independence standards whole mind-set if you're willing this no 1 will try to write you with advertising money or free flights showcase something instead you can if so inclined
start following only sell freezing artists or you just click we made up diet of music released by independent labels only maybe through all this you can even emulate something that would cause the immense moments of music consumption that were often part of the experience of until 15 years ago at that time the actual type of music you listen to was an important and in order to be able to listen to this music usually have engage with people go to places and spend your money off the mainstream then again everyone loves a good hit and if that currently might be for most of us maybe not constant lacking than who cares of its major company that turns on the dollars with personally I'm more interested in each streams and in the mainstream and the
current ecology of music the new politics of taste making allow for great opportunities for spreading all sorts of minority esthetics in each sounds to use a term from the discussion
about the so called alter series like breaking Bad of the wire that would never have impressive ratings when the premiere on network TV but find a much larger audience the secondary channels like DVDs and so forth there are now ways of quite elaborate ways actually of distribution you could call narrowcasting as opposed to growth broadcasting and this is for me 1 of the key aspects of today's curated models instead of opinion leading individuals and collectives that primarily determine consensus of supposed to fit for all we know the multitude of collectively let opinions that allow for more diversity and this makes it possible that more different kinds of music find our attention even if less single-track will find the attention of all of you raters or influences or whatever we call them play a vital role in terms of charging on music with meaning even if it's now more of a social meaning then President and expressive and if you look at music television you could say that it has benefited musicians were able to turn into visual artists well on the other hand with the Internet arrived at 1st of all turn music into audio files so now everybody has to see how we can make more of them and besides various strategies to charge music with more meaning by way of visuals videos artist logs or performance aspects and so forth the rated music recommendation recommendation can add a social factor to this node it's not necessarily always only the music itself that determines what we like it can also be who else is listening to this music and was talking about it and how it is released so in the final analysis it is probably the relevant on which platform you receive your recommendations of maybe specific platforms would influence your influences so now we should probably look at some of those platforms and to do this I'm happy to have balked at my side to prepare this little thing which I don't know that works but when we haven't had OK so we will have a bit
more visual now and start talking about some of the later developments that we might be interested to hear about yeah yeah a it and I'm introducing local Quebec list is all you want to introduce yourself you had a nice time
OK is managing director of fine genes which is the digital distribution service company find expressing that correctly based in Hamburg which ourselves a lot of labels and I artists to bring their music into the general market and also he's on the board of the beauty the German independent label association and at my languages the bigger brother of that on international level so I don't look for it you don't want to it the
all and it it also the things that I it is if the the OK and no everyone nice firing
of here and now and we present a um I've made a little presentation on kind of some buzz words we're trying to reach out to what is happening to music consumption in um uh related to uh what is curation what is influencing what is uh consuming music analysis changing the landscape and there we can definitely say of course not only because of streaming services but
with everything what is happening at the moment we see different kinds of curation all models are coming and these are helping of course artists and especially some the shot is depending on the curation my on the curation model in being more visible in the landscape because what we're facing in in Germany usually is that and mainstream media usually has the um kind of market shelf 6 % of niche content even tho the market share for um music sold often use content is around like 25 % of the independent related um music as seen and with the help of curation all
models and we can see the development from the US of course coming with Pandora with uh mother streaming services they had these kind of numbers of reach out of independent and music is kind of getting and not even up to the 25 % I was just talking about is just even going much more higher up to 40 % where of course and established artists are being less visible in certain curation of system which which we want to talk about now so that it had prepared 1 question that that was directed and the sort of so with the idea that she rated models might help independent models were unknown artist better than uh algorithms which might relate to sort of our of something that's called the filter bubble as well that strong things usually enforced by something like algorithms I'm sure the 2nd part of the music recommendation session will be having some ideas about that and you actually had uh some market research holds that about it recently that basically said yes it is sort of a more important to be in editorial systems which are the rate in terms of streaming services were talking about now and rather than the algorithm based stuff to them on the on the edges we can definitely see this slide off
where editorial based systems like for example iTunes or even editorial based streaming services like um in um and others where the release gets presented and
and to the audience is of course much more interesting for newcomer bands because then they get a visibility which they usually not getting when it's get getting more to the human and social factors so sharing and then we see lots of of course of already known had that it's like people are listening to music they they known for years and 8 which is quite common so it is really then interesting that a mixture between the human and social and of course the uhm uhm user based sharing neat is needed for getting both um forgetting both effects for newcomers and for an established bands um it is much easier as we all can see of course with the Daft Punk example where of course everybody's listening to it whether it's a 50 year old um uh more than 70 yield the or 7 yields the
1 thing that I came across in the in the preparation for the session was a study done by Nielsen music which is sort of the more industry related to the organization is understand who were looking at the discovery of music and the conversion of discovery
into actual consumption meaning in that regard for being in buying of downloads or the and um they still find the direct recommendation by radio is very important that's not so much what we talk about but also the recommendations by friends to be very effective which relates to sort of the I think the relation of trust that is necessary to have do we give the the final impact for somebody to say OK I'll I'll give it my attention not because the the starting page of my streaming services so but because my friend all those always had the best new music dish has recommended it and so we tried and find out as I understand they didn't really have figures for that how important for example insights 45 it is to have a social recommendation that I think in general there is a trend that this is a huge set of factor in consumption is the possibility to get I did not everybody in Europe friends whatever lists you have and sort of sorted by way of their music preferences and and in that regard to this like to services that will probably soon rule out we could look at it for a moment before actually been in Austin I think presented by speech the Communist new venture which employed you Rogers of crops and sort of figure it for their new music service which basically is going to be you know what I mean by this is another and this because they just bought 1 streaming Service which then of course has a kind of new expect like accu rated model where they are really putting together more than 1 thousand and influences who are preparing play list but this is especially when I saw it it was nothing else than which we can already see on such several other streaming services is already being done by for example Spotify creating artist latest looking for it influences whether they are actors or they are and musicians and they are preparing all the lists which now of course pizzas trying to somehow combine with the brand they already have with that phone Dr. beats of of Dr. Drake and this is of course where professional curation and this kind of cross curation always speak about cross-marketing but it's a bit more like cross curation get me some testimonials which are like in my product in this case the headphones they are also presented on the streaming service and therefore people might get more attached and more easy to listen to some music because I've always wanted to know what Dr. Draves listening to but I am not too convinced of this kind of model at the moment well it seems like from early mining heavy investment scheme ways by a lot of of endorsements so not nothing you technically maybe Twitter music the other sort of as of now only available in the English language countries it's understand service might have more of an impact adenyl over the did we have any sequential for that you know we don't have we don't have Twitter's is lounge has announced in the US and UK and some other countries have as a service where you can directly access music recommendations as far as there are done
on you tube sporty finite use I believe directly embedded into a Twitter environment where you can see what your friends are posting from this name channels or get the usual kind of popular now with analysis of of traffic in that direction I believe that's going to be quite influential but as of now not available in Germany so we might look at some or all of them that kind of new services which upcoming in the um English-speaking countries are coming here very late for several reasons mostly because of licensing reasons and this is of course there are some are to be seen very crucial because like newcomers
and artists which upcoming who want to to in Germany all i young bands which want to get known via Turing and I'm not really represented in those kind of services because they're just not available in Germany and this is definitely something where young and newcomer um bands in the other countries do have it much more easy way it's much more easy for them in other countries than in Germany so maybe we should
also talk about another service so far not available you well we on that topic that would be a turntable F and I think there's a 2 was just about 1 of the images on the screen on I
learned yesterday that this is an instance of gain deification that is at work there besides other stuff Germany can now says sounds drop at on Spotify that has some aspects of this in in anatomy would like to explain how to that work I mean it's just like presenting regulates even in the in mixed way to audience which can go into your room in which can listen to different DJs in this kind of um surrounding everybody can open up the room and everybody can open up a rope invite people or people can't even if you make the room public can go into the room and can this and somebody can play than rackets and even more deejays can presented Reckitt so there are like 5 up to 5 people being able to present rackets to their friends so it's a bit like a meeting of the Keller room with old vinyl records and everybody's playing a vine let's is recommending something to the other something that you love you love having examined half yourself and on and the publicity can then say 0 I like to track i don't like to track of course there are possibilities and you just can't listen to the stream more you are able of course to download it somewhere at some source to a certain degree I
find this kind of model not ultimately desirable because it's a employs sort of extra effects to get attention for music we talked about the sort of an economy of attention which is today very important that people not only look at the financial gains immediate ones but 1st of all the attention which are makes it sort of this very open to all concepts that are employing nonmusical on specific musical strategies like things have to be spectacular very sort of gave me and all that which is not suitable for all
kinds of music so I think that's always something wherever have the need for destruction is getting bigger and all this so so I prefer basically uh other aspects I guess of of PureTalk models which we getting there 2 minutes I'm out of time we should maybe shortly mention shuffler and because it is here which is the from each of the vendors were very nice because it's more often um kind of source well all blocks in um that aggregated into 1 source in different and we sold on the screen in different and uh um insurers and everybody can listen up to the John Ritter you feel very interested in and you can listen to new music which has been put on some blogs but this is of course something where you're not q rated by friends in this case it's curation by an all blocks of the world and it is by popular by popularity so that and the more 1 trick got posted into blocks the more and visibility is it in shuffle less and it's a kind of hype machine and in the same like of the model may be looking a bit a more interesting and is a getting the music from different sources as well of course using tube as a kind of source using um um not only um um SoundCloud all all um in In stream possibilities as I understand for you you itself is also quite important as an income stream out already for music or that doesn't work that the American determinized from that definite definitely it is and I mean what we really have to see if the kind of we see that the user is getting more and more of a distributor which is why I call that the kind of discrete views which we need to a certain extent you
get people listening to music they are not that I'm aware of yet and this is of course with the help of all the different streaming models with the
help of different uh and um even L. L. U. rhythmic Q rated models you can see what has been shared what what is flowing in the net and by popularity or by human interaction you can see different kind of music which is definitely a big fear for public broadcasters especially in the radio um um scenario because the radio is so um programmed in the
stiff way that of course you just listening to tests to to program kind of play lists which in the end not to rated at all anymore and this is a fight unless you're listening to bite them interesting is working abideth in which is a nice radio and channel by the way uh and um but therefore we see the kind of really Captain bag of Q rated models
which have been in the past the radio stations and the radio stations because I talked to some public broadcasters in germany at just recently they are not really being aware of becoming totally useless and maybe a few years OK so I think we have to stop here and the 4th thank you so sorry I figure much as like a 15 minute prelude you going to announce the rate for the next session thank you can sample can
M. uh
and the fact that