Niche vs Mainstream
Anas Buhayh discusses multi-stakeholder fairness in recommender systems and the S'mores framework—a simulation allowing users to choose between mainstream and niche algorithms. His research shows specialized recommenders improve utility for niche users while raising questions about filter bubbles and data privacy.
Guest
Anas Buhayh: Anas Buhayh is a PhD student at the Recommender Systems Lab at the University of Colorado Boulder, where his research investigates alternative architectures and technologies for building responsible and governable AI systems. His work emphasizes the importance of addressing the needs of stakeholders in sociotechnical systems. Anas also volunteers as a consultant for recommender system design and development. A Fulbright alumnus from Libya, Anas holds a master’s degree in Information Management from Central Michigan University and a bachelor’s degree in Automation from the College of Computer Technologies in Tripoli, Libya. His professional experience includes roles in monitoring and evaluation with the World Health Organization and the International Rescue Committee, where he focused on aligning project outcomes with stakeholder needs. Anas is passionate about data analytics, machine learning, and systems design. In his free time, he enjoys programming, playing the guitar, and staying active.
Transcript
Kyle Polich : Welcome to Data Skeptic a podcast exploring the methods use cases and consequences of recommender systems
Kyle Polich : Welcome to another installment of Data skeptic Recomer Systems Today we're gonna start entertaining the idea of decoupled recommender systems What do I mean by that Today pretty much any recommender system you're going to encounter is an in-house thing If you go to Netflix the recommendations you get there are from the algorithm produced by people who work at Netflix
Kyle Polich : But what if there could be a marketplace there the way your phone has an app store there could be an algorithm store
Kyle Polich : Of course that would require these platforms to open up but perhaps there's a very good reason they should do so Towards that end today we're going to discuss the simulator for modular Recommendation ecosystem or S'mores framework We get deeper into these decoupling ideas and explore how s'mores can be an empirical tool for studying how ideas might play out in that space
Anna Bohai : My name is Anna Bohai I am a student at the University of Colorado Boulder provised by Doctor Robin Burke I'm at the Information Science department at that recommender systems lab and most of our work in this lab focuses around multi-stakeholder fairness and recommender systems Yeah and basically the work I do is mainly around evaluating alternative recommender systems
Kyle Polich : Use the term multi-stakeholder could we talk about who those stakeholders are
Anna Bohai : Yeah so usually whenever we think about a multi-stakeholder recommender system we think about recommender systems that work as matchmakers So think about having someone who's providing a service So for instance if we're talking about Spotify that can be the artist or think about a YouTube that can be like a content creator DoorDash the restaurant or the driver
Anna Bohai : Someone who's receiving the service So the one who's providing the service we call them providers and the person who's receiving the service we call them consumers And that's the user that we usually call that the user uh on an everyday basis And that's the person who's like listening to the music making the order on DoorDash We have the platform which is something that we cannot ignore because there's a lot of conversation like oh
Anna Bohai : We need to do the matchmaking but we forget that the the platform also needs to you know like make money to keep running These are the three stakeholders we talk about And most of the time when you see in the literature when whenever we hear a user we don't think of providers as users for some reason but they are as valuable uh users in the uh in this whole ecosystem
Kyle Polich : So when you consider fairness is it for the provider the consumer or the platform or maybe all of the above
Anna Bohai : All of the above yeah you need to think about everyone on the platform because recommendation as an ecosystem needs everybody to be happy to some extent to stay on the platform So let's say if I joined the platform as a provider let's say a content
Anna Bohai : Creator and my content is not reaching anybody then I'm going to leave that platform And then you would have like a platform with only like 3 or 4 content creators which is like make the content So like how do you balance between all of these multi-objectives on the platform
Kyle Polich : What would it mean for a recommender system to be unfair
Anna Bohai : So the the fairness issues in recommender systems are similar to like machine learning as a whole So you have two fairness issues that you would usually find in the literature which are representative fairness and allocated fairness So let's start with the first one Let's say representative fairness and talk about it in the
Anna Bohai : The idea of like recommender system So let's say you have a job recommendations and we're recommending to females uh like a less paying jobs or like a specific stereotype of jobs and you're presenting to males a different kind of job So that's as a as a receiving end Also another way it could be like I would
Anna Bohai : Jump on a search image and they look up I don't know an Arab man or like an Arab man So like what kind of photo would you see whenever you look up that Because that is in itself is a retrieval so it's a recommendation So that's the representative part But the other part is the allocated fairness which is let's say we're talking about like
Anna Bohai : Content creators for instance So the idea here me as a new content creator joining the platform will I get any chance to be recommended or only like the very famous people on that platform will be presented to the consumers So these are kind of like the two main things we talk about whenever we talk about fairness and machine learning
Kyle Polich : Well I know you probably can't have surveyed all of industry and all of implementations There's a lot of recommender systems out there but I assume most of these are written in-house Is that true As like
Anna Bohai : the recommender systems Yeah
Kyle Polich : so like at Spotify or YouTube whatever they're doing it's probably a trade secret by employees inside the company
Anna Bohai : Yeah exactly I mean the thing about recommender systems as a whole they are multi-stage So there are like two narratives Uh the first narrative people think about recommender system as just like one algorithm You just like put that algorithm and then you get a specific output But in reality that it's multi-stages So for instance we're talking about an enormous platform like YouTube you cannot
Anna Bohai : Put every item like those millions of videos or maybe I don't know billions at this point you would put them and uh the algorithm and they would spit out recommendations for a specific user because that's that's so many So you have different stages So you have the retrieval which is you would get a
Anna Bohai : A portion of those items and then you would put them in an algorithm and then that algorithm would focus on the relevance Let's say we picked 1000 items we rank them by the relevance to the user And then we do another stage which is the uh filtering So we filter whatever items that we don't want to show to the users and then we do re-ranking So most of the
Anna Bohai : Fairness happens at the last re-ranking stage However if any let's say bias or unfairness happens in the previous stages of course whatever you fix at the end is not going to fix it So let's say if I'm not picking the niche user from the beginning and my like 1st 1000 item that niche user is not going to make it to the end whenever I do like some sort of re-ranking
Anna Bohai : And that's where you see like most of the responsible AI work is happening in the re-ranking phase because that's when you have the opportunity of like OK let me add users from this specific group let me put these users more on top of the list or something like that
Kyle Polich : Re-ranking makes sense it's intuitive Are there other ways to approach the challenge
Anna Bohai : That is where we come with our idea So we were like OK we we have the the re-ranking but what if we just make a new recommender system What if we let people create more like algorithm developers have their own algorithms that they would attach to this platform and then you would have
Anna Bohai : Like a curated algorithm and by this plurality of algorithms by having like more algorithms we might curate to different users and also different providers may make it to the consumer at the end
Kyle Polich : Is that the s'mores framework we started talking about
Anna Bohai : Yeah the the idea there comes from the uh the algorithm store The algorithm store have like two people started talking about it There's like this manifesto which is called like the three-legged stool
Anna Bohai : And the idea comes there from like what if for instance on a platform like Mastodon like a federated social media platform what if you let users create their own algorithms and then they would create their own needs So you would have like for instance a user would choose from like a spectrum of algorithms you're like oh I want this one that for instance
Anna Bohai : gives content about this specific niche sport and I don't want my niche sport to be within the like this big mega sports algorithm
Anna Bohai : So I'm going to choose that one And also you can think about it as like uh an algorithm that fits my values So if I am on DoorDash I might want an algorithm with the restaurants with only like ethical meats or like kosher meat or something like that So whenever we have this idea of like algorithm pluralism we would have more opportunities for users to choose
Anna Bohai : Use the path that they want And there's this another idea by Fukuyama which is the middleware that is basically just the idea of like middleware and software engineering So you would have this software that would work alongside the system and it is more of like a modular component where you can replace one algorithm with another algorithm
Kyle Polich : So this idea of the friendly neighborhood algorithm store is interesting it's the opportunity for researchers like yourself to come up with a new idea and put it out there within the platform I don't know do like a multi-arm bandit thing or or pick their favorites and experiment with them or is your vision more that the user has a choice here
Anna Bohai : So we see it as the user has a choice The closest example here is Blue Sky The Blue Sky has this idea which is I think started in like in the past couple of years uh which is the customized feeds So you can think about it the same way So the user would try to look up a specific feed or an algorithm There should be like specific metadata about that algorithm and how it works And then the user can choose
Anna Bohai : their own algorithm Because what we're trying to do is to lessen the control of the main platform that owns the item let's say We're trying to listen to control of like what's trying to surface and give that control to the algorithm designer because we don't want this centralized power of like only this platform can decide what the user sees We want more people to decide
Kyle Polich : I'm curious about you know being able to have that choice be empowered to pick my algorithm but it's also another step I have to take So what are you actually trying to measure in your experiments
Anna Bohai : That's a really good question and that is the the question of the the user burden So like whenever we we add more algorithms we're asking the users to investigate and see their own algorithm
Anna Bohai : And like find the one that serves them And that's and we try to see it more of like a a subscription model So like for instance uh you choose whatever newspaper that you want to read or like whatever uh streaming service that you want to use And it's
Anna Bohai : This same idea So we started with this research question OK if we have multiple algorithms instead of one would that be beneficial to a specific group of users and or like the the users who are interested in that specific content
Anna Bohai : Because there's no platform out there to try these ideas we had to build our own environment And we were like OK what is the best way to do this And that's when we came up with the idea of what if we just make a simulation environment So we're just going to simulate this whole
Anna Bohai : Basically ecosystem where we have providers and we have the platforms and we have the consumers So basically we use real data So there are like multiple data sets uh out there that recommender system researchers machine learning researchers use So for instance one of the famous ones is called MovieLens and that is user ratings data set like so basically user rated a lot of movies
Anna Bohai : And what we did there is that we took all of the movies like that are on the data set We found their studios from a different data set and then we were like OK these are the providers So the provider is the studio in this ecosystem And then we already had the users so we used those users
Anna Bohai : And then the last part which is the platform And we're we basically because we're thinking about this idea of like the algorithm store we had to simplify it so we can measure what is that change So we just created two recommenders So like the store has 2 recommenders at this point
Anna Bohai : So it has one mainstream recommender and one niche recommender The niche recommender basically focuses on a specific genre of movies and the mainstream would just see whatever like other users on the platform wants to see So we basically we use an algorithm called SPD
Anna Bohai : Yeah basically it's just like uh one of the the popular recommendation algorithms We ran the experiment and we had a multiple switching mechanisms So we had a utility uh measurement which basically says if the user does not see the genres that they want to see in a recommendation list then the utility would decrease
Anna Bohai : And if they see the genres that they want to see in that recommendation list then the utility would increase And we had a specific threshold and we said if the user's utility would drop below that threshold then the user would go and choose a different uh recommender So we had this idea of like a switching mechanism
Anna Bohai : By the end of the experiment we found that users who moved to the recommender that tailored more to their needs would have a higher utility and also the providers who did not make it in the mainstream recommender because they were a niche provider
Anna Bohai : When you have a a recommender that focuses on their content then it's really easy for those users to find that content instead of trying to find it in this abundance of content on the mainstream recommender
Kyle Polich : How many logins does it take to answer one business question ERP for orders CRM for customers HR for headcount accounting for spend then someone sends you a spreadsheet that contradicts all of those other systems
Kyle Polich : Scalar ends that chaos It unifies every business system structured and unstructured data into one governing source of truth Ask questions in plain English Get instant accurate answers No exports No chasing people across tools
Kyle Polich : No more 3 plus platform treasure hunts Just to close a simple report Scaler works with Snowflake Data bricks and every other data provider you already trust No migrations are required It understands what your data means where it lives and gives you access 10x faster saving you time and money Go to scaler.com to get your free demo today That's S C A Y L O R.com
Kyle Polich : Once again Scaler S C A Y L O R dot com Get a free demo of Scaler and find and finally unify your business data once and for all
Kyle Polich : So you mentioned having two recommenders obviously the one's for mainstream and then one is niche You said for a specific genre was there one genre you picked or did you have several niche genre models
Anna Bohai : So we tried with different data sets for the paper that we published we had the horror for a movie lens so like the horror genre
Anna Bohai : And for we use an Abar data set which is a music data set and for that one we use Soul Funk For each data set we tried one niche Ideally in a simulation environment and that's the work that we're working on right now you don't want just two You want this idea of niche to expand So let's say users who go below a threshold and then you try to create a niche from that group of users below the threshold and then
Anna Bohai : Until you reach 10 amount of recommenders So having these strict simulation environments are really good in a sense of like they would give you the like the first picture like oh this is a potential thing But I think there's a lot of work needs to be done to come to the idea of like oh this is the thing that we need to do And these are like the clear trade-offs of this work
Kyle Polich : Well 2 different recommenders can differ in their training data in their specific algorithm or maybe just some parameters to the algorithm How did your models differ
Anna Bohai : For both of the recommenders we want it to be consistent in terms of the parameters However the data is what's different So the niche recommender only recommends items that would have the genre that we defined So let's say if it's we call it like the horror recommender then we'd only recommend movies that has the horror genre The mainstream would basically recommend everything
Anna Bohai : The data is also split So as a simulation environment the users would arrive so let's say like the first day of the simulation that's where the users are going to interact with the recommender system But before that the recommender system does not know anything about the user So we're trying to simulate an environment that the recommender sees every user as a cold start case So that's a new user
Anna Bohai : So because we start with the mainstream recommender then the mainstream recommender would build profiles on the users and then we'd use the data that the user selected as a training data for that user And then when the user moves to the other recommender the niche recommender does not know anything about the user at that point because it's also facing the same problem if it's a cold start The
Anna Bohai : This is a new user that is arriving So we have this separation of data between the two recommenders And that is a question that we're also interested in which is should users transfer their data Like if I'm moving from one recommender to the other recommender should I take my data with me And there is a research
Anna Bohai : That is evolving from the GDPR and some other legislations which is the idea of data portability So me as a person who's moving between recommenders should I take my data with me and how would that be even beneficial to the other platform if I take my data with me and what are also like the trade-offs there
Kyle Polich : In your simulation the users are gonna follow that switching mechanism you described earlier Do you have any high-level observations about like how often they switched or you know things along those lines
Kyle Polich : Yeah
Anna Bohai : so
Anna Bohai : the users or we call them like the niche users which are the users who are interested in like the horror genre for movieence or the soul funk for the umpire
Anna Bohai : Those users I would say like 90% of them would switch to the recommender really early on like the ni recommender and then only a few of them would keep switching back and forth And we had this idea of like back and forth switching because some users would not find any of these two recommenders
Anna Bohai : Satisfying So let's say we can group the users to like 33 groups We have the people who have like the action and the drama Those people are happy Like and they're mainstream recommended they're always going to get the the action and drama There's a an abundance there And we also have our niche recommenders who are like really interested in the horror So those would move
Anna Bohai : The the niche recommender and they will be happy there And then we have a group of users who would not be happy in these two So these this group of users will be like switching to the generic recommender They would get good recommendations then it's going to get lower then they're going to switch to the other one And that's when you need a third recommender to kind of like solve this
Anna Bohai : Problem But uh you don't want to keep increasing recommenders until every user will get their own recommender right So there is like a threshold there that you need to find where OK until what point I can add recommenders while I'm adding value to the users Like at what point does that stop adding value let's say
Kyle Polich : Well the more traditional approach is just the monolithic recommender it would have the mainstream one and that's it How do you measure the improvement you get by adding in the option for the niche recommender
Anna Bohai : Yeah so we basically measure the utility for every provider So for the providers we measure their utility for the generic providers and then we measure the utility for the niche providers And then for the consumers we
Anna Bohai : The same thing We measure for the the generic and the niche And what we found that for the generic providers they would use some utility because basically some of the like the views or the clicks that they were getting they were getting from niche consumers And now those clicks are moving somewhere else So like those niche consumers are going to click on like the niche providers
Anna Bohai : For the niche providers they basically gained so much utility because they have a recommender for themselves and users Whenever they go there they'll be like oh this is like like the niche content that I'm interested in So every niche provider will get some sort of uh exposure on that recommender As for the uh consumers we also sell the same thing So like niche niche consumers would have a higher utility However
Anna Bohai : Then the generic consumers sometimes when they switch to the niche recommender they only see stuff that they're not interested in So if if the consumer makes the switch to the other recommender then their utility might drop But we also found some users who we did not label as niche but they stayed on the niche recommender because they liked it enough to stay there or like they liked it more than the mainstream one
Anna Bohai : So there's definitely a trade-off to make there And for the provider trade-off I can argue that for those mainstream providers like those clicks weren't supposed to be yours anyway so they should go like to the niche But for the consumer side that's the one that you don't want You don't want uh if I'm like interested in like comedy uh don't move me to a a horror recommender That does not make any sense
Kyle Polich : It's intuitive to me that a niche you know horror lover who gets the niche recommender is more likely to enjoy it And I see your point about there's a marginal cost those views went away from maybe mainstream providers If there were a big platform let's say it's YouTube or Spotify or whoever who was interested in this idea but the product manager says well I can't afford those costs how do you look at it practically
Anna Bohai : Yeah that's a that's
Anna Bohai : a really good question So I think because of the simulation environment that we have we are facing those costs I don't think that those costs are going to translate
Anna Bohai : Directly to like those platforms Because one thing for instance that we did not account for for our stimulation environment is the say like the name of that recommender So for us as the creators we know that that's a horror niche recommender But in the the sweatshir
Anna Bohai : I think that the user does not know that it's a a horror recommender They just think of it's like Oh this is a different recommender that uh that I need to check out So let's say if we're thinking about a platform like YouTube I like playing guitar so I'm I'm interested in like guitar content
Anna Bohai : So if someone created a a recommender for guitar tutorials or like more specific like guitar tutorials with tabs or something like that then I as a user I know that this recommender is for this specific thing
Anna Bohai : So that kind of like lowers this cost of the user seeing stuff that they don't like And I would argue that me as someone who's like going to YouTube and seeing like everything on the same page that's when I'm losing utility Uh and I would gain it more if I can choose what I want to see actually
Kyle Polich : Are there any technical barriers to getting an idea like this that there are third-party algorithms on a platform Are there technical barriers to making that a reality
Anna Bohai : There are some side effects of our recommended algorithm So for instance we have the filter bubbles where everybody's seeing the same thing
Anna Bohai : So like for instance if you are either on the uh on social media and you're either like on the right or the left you would only see content that is from that political spectrum And that's what we call like a filter bubble If you let users design their own algorithms they might even exacerbate like that idea of like oh I'm just going to focus on
Anna Bohai : This specific content I'm not going to allow for any content from like the opposing party or like opposing ideology or even like slightly outside of that ideology
Anna Bohai : When uh Fukuyama introduced the idea of middleware he said that a platform oversight might be helpful there So the the platform itself need to have a specific requirement for someone to create an algorithm Another thing can be like a transparency
Anna Bohai : So we need to have more transparent algorithm development Like can the users for instance see this third party algorithm and like how it's developed And not everybody's going to going to understand like the technical part of that but we know it's open for anybody who is interested As for like the technical development side
Anna Bohai : Because we see this idea of as like the platform has the storage of the content and these third party algorithms just have access to that storage so like
Anna Bohai : Development wise it's not a very hard idea to develop It's more about like all of these trade-offs that you need to consider Like as we said like the user privacy like is every third party algorithm going to own the data Is the main platform owning the data and then these third party would have like
Anna Bohai : A specific access So there are a lot of things there that we like a platform needs to think about to make this happen
Kyle Polich : So what is the incentive for a platform to adopt third-party algorithms
Anna Bohai : I would say the main incentive would be that users would have a better experience on the platform as they would find the content that they would actually be interested in Because I think there's also like differing kinds of platforms There are there are the platforms who focuses mainly on let me just like put more ads in front of the users I want to surface specific material in front of the users For those platforms probably this is not what they're looking for But with the platform
Anna Bohai : Platforms do actually want the user to have a good experience on their platform and they're not just trying to like go over engagement and these kind of things Those platforms basically can benefit because they are opening a new market where
Anna Bohai : Other ideas can come in and then those ideas might be a bit more beneficial to the users So again like back to this idea of me like in guitar so a recommender that focuses only on that that would be very helpful for me because like when I think about YouTube there is what I want to see and that is like what the recommender is giving me So what I want to see is more of the things that are
Anna Bohai : Aligned with my interests but the recommender is like oh you like this cat video so you're only getting cat videos from now on So this kind of constraints of a third party algorithm can help the user also choose what they want Then the user would be happier also being on a specific platform
Kyle Polich : From your perspective is there a trend in industry to move in this direction or are we maybe just at the talking about it phase
Anna Bohai : I think because Blue Sky had something similar so at least we know that there is one part of the industry that is aiming at like going in that direction and it's very promising because they have this idea of like created feeds where users can create their own feed and add specific filters and then you have a feed basically uh with a specific sort of content And I think also there are
Anna Bohai : Some like startups that are
Anna Bohai : Focusing on this idea with like building feeds for Blue Sky Um I don't remember the name of the the one the very famous one but basically there are some companies who
Anna Bohai : are helping users who are interested in creating their own feeds I think Spotify when they added this idea of like you create your own playlist and even like YouTube and then you can share your playlist that is one version of it That is one version of me like creating tailoring content to other users I would argue that Spotify's AI prompted stuff is like taken over but a lot of like YouTube playlists are still very popular I would say
Kyle Polich : So the s'mores framework is available on your GitHub page people can test it out or extend it or uh who knows what else Mhm Tell me a little bit about that choice and about what users will find there
Anna Bohai : Yeah so we uh we've been working on this simulation environment for almost like 3 years now and just like tweaking stuff adding stuff It is on GitHub on our like that recommended systems lab repo And once you go there you can just clone it There's information there about how to install it
Anna Bohai : We have also other repos for experiments So if you want to see an example of uh how to run an experiment we we we also published the uh the experiment repos because we we are separating of like this is the simulation environment and we have other repos for simulations like the experimentations that we use for our papers and those usually would have like our results and our inputs
Anna Bohai : Yeah it is it is available and we did not talk about like contributions yet but I'm sure that will happen at a specific point
Kyle Polich : Very cool yeah Well what's next for you
Anna Bohai : So I am really interested in this idea of alternative recommender systems and user choice The next study that I am interested in is OK so we gave users their own algorithms Let's say we created this niche algorithm and users joined that niche algorithm
Anna Bohai : How can we allow those users on the platform control what is on the algorithm So we we go more to the another form of user agency on the the re-ranking stage and I'm mainly thinking about it as recommender systems uh like the the group recommender lens So if you have
Anna Bohai : Let's say even like a Mastodon instance with 1000 or 2000 users how would those users decide what their recommender is going to look like or like if you have a subreddit how would those people on your subreddit if if they have the chance how would they choose what's allowed on their feed uh how it's ranked what is important to them So moving more to the uh the agency part
Kyle Polich : Makes sense yeah And is there anywhere listeners can follow you online
Anna Bohai : I'm usually active on LinkedIn so you can type my name Anna Bohai on LinkedIn and I don't know if there's anyone out there with the same spelling of my name so you're
Kyle Polich : probably
Anna Bohai : going
Kyle Polich : to find me lucky in that regard yeah Well thank you so much for taking the time to come on and share your work
Anna Bohai : Absolutely thank
Anna Bohai : you Kyle Thank you for having me