My reflections on the first two weeks of my masters module on Data Information and Technologies, how our data is used to curate what we see and the roles and responsibilities of Library and Information Science (LIS).
“It’s better for me, better for you, it’s better for them”
In our increasingly digital world we have a lot more platforms vying for our attention, time and money. One way that these digital platforms attempt to do this is by using our data to tailor and curate what information we see, be it search engine results, suggested videos on YouTube or what titles are displayed for us on Netflix.
It’s easy to see why, with a platform like Netflix with nearly 14,000 titles (Cook, 2020) to choose from or on YouTube where over 500 hours of content is uploaded every minute (Smith, 2020), some form of curation is necessary to help us find what we want and to avoid information overload. This is something that was also recognised by Floridi, when he said, Customisation means having the world adapting to what we need… And recommendations based on our interests are better than random ones (Floridi, 2016).
Evidently, these recommendations are not just helpful for users, over 80% of titles watched on Netflix are discovered through their recommendation system, but also crucial for platforms, Netflix report that they only have 90 seconds to capture our attention or we’ll move onto another activity (Blattmann, 2018).
“We fear change”
However, in expecting algorithm curation based on our current tastes (or at least what digital platforms can gauge based on the data they have on us) we are potentially closing ourselves off to discovering new information, hearing diverse voices, or perhaps just an opinion that is different to our own, as Floridi put it: technologies… can mummify so effectively who we are and can be. (Floridi, 2016)
I have noticed this with my Netflix recommendations, I watch a lot of horror films on Netflix, so this tends to be the kind of programming it then recommends for me to watch next. Great, of course I want to watch the cinematic classic Wishmaster. But I do have quite an eclectic taste in films (humble brag) that wasn’t being represented by my recommendations. So, as an experiment, I clicked on my sister’s profile (who hates horror) to see what recommendations she was getting and to my surprise it was like entering a whole new world, with movies and TV shows I didn’t even know were on the platform. How to Lose a Guy in 10 Days? Watch now!
My reliance on their curation-algorithm and their need to get my attention quickly, meant that I was not exploring the full extent of their library and potentially getting siloed into a particular interest group, making me fall into a concern of Floridi’s that digital technologies may move from being able to spot who we are to actually making sure that we become who they say we are, and do not change. (Floridi, 2016)
“Trust – who do you?”
Furthermore, a limitation with curation algorithms is that many digital platforms depend on ad revenue, so they can then make information harder to find for people who need it because it’s getting clicks from the wrong (or unintended) audience. This is outlined further in an interesting video from Stacks & Facts on the limitations of algorithms when deciding on what content to show people, where he uses the example of a video that is intended for women being watched almost exclusively by men because of these factors.
He also cites some ways in which librarians already overcome some of these challenges in information discovery. That our humanity allows us to have an understanding and make connections that algorithms cannot and also through the practice of more in-depth and accurate labelling of information a user can more effectively determine the relevance of a piece of information, rather than just titles and images that are designed to entice them to click. This brings me nicely onto…
The role of LIS
Information and library sciences are, I think, by definition, heavily involved in how digital platforms present information to us. If we consider the role of LIS to be involved with all aspects of the communication chain, the creation, organization, management, communication and use of recorded information (Bawden and Robinson), to again take Netflix as an example, this would include the way in which information is organised, how it can be retrieved by users, analysing how people use the platform and what value they get from it.
Furthermore, Netflix’s recommendation algorithm is powered by making connections based on metadata, this metadata is created by information professionals categorising every aspect of their library (Blattmann, 2018) to allow the algorithm to make more sophisticated, or in some cases, tenuous links between films and television programmes. Ah yes, my favourite genre of film: Violent Nightmare-Vacation Movies.
However, as I have outlined above, despite some clear advantages in this perhaps information-saturated world, there are limitations with relying solely on algorithm-curated information. I think this is where LIS professionals can also step in, with their mandate to promote information literacy (CILIP, 2018) and experience in data literacy (Moran, 2019), to help equip people with the ability to effectively find the information that they need and to help people understand what data is taken from them, how it is used and how their actions can then be affected or manipulated by digital platforms.
So don’t give in to apathy and accept the constraints of a digital world an algorithm curates for you, your How to Lose a Guy in 10 Days is out there, find it.
Cook, S. (29/07/2020) 50+ Netflix statistics and facts stats that define the company’s dominance. Available at https://www.comparitech.com/blog/vpn-privacy/netflix-statistics-facts-figures/ (Accessed: 07/10/2020)
Smith, K. (21/02/2020) 57 Fascinating and Incredible YouTube Statistics. Available at https://www.brandwatch.com/blog/youtube-stats/ (Accessed: 07/10/2020)
Floridi, L. (24/04/2016) The self-fulfilling prophesy. Available at: https://www.schirn.de/en/magazine/context/the_self_fulfilling_prophesy/ (Accessed: 07/10/2020)
Blattmann, J. (02/08/2018) Netflix: Binging on the Algorithm. Available at: https://uxplanet.org/netflix-binging-on-the-algorithm-a3a74a6c1f59 (Accessed: 07/10/2020)
Stacks & Facts (2019) YouTube’s Algorithms (and Masturbation). Available at https://www.youtube.com/watch?v=dXHzEj1cVFs (Accessed: 07/10/2020)
Bawden, D & Robinson, L. (no date) Library and Information Science. Available at https://hcommons.org/deposits/objects/hc:12808/datastreams/CONTENT/content (Accessed: 07/10/2020)
CILIP (2018) CILIP Definition of Information Literacy. Available at https://infolit.org.uk/ILdefinitionCILIP2018.pdf (Accessed: 07/10/2020)
Moran, G. (31/08/2019) We’re in a Data Literacy Crisis. Could Librarians Be the Superheroes We Need? Available at: https://fortune.com/2019/08/31/data-literacy-crisis-librarians-library/ (Accessed: 07/10/2020)