Browsed by
Tag: Tumblr

Exploring Algorithms, part 2

Exploring Algorithms, part 2

More algorithmic exploration

Netflix

While fooling around on Netflix, I thought I would see what would happen if I chose to watch a film from the ‘horror’ genre – of which I would never usually choose as I detest this genre!

The first movie in this genre on the list was: The Last Days On Mars, which I watched a portion of and was relieved to discover it was not that scary. Because I chose a film that was out of my normal realm of preferred viewing (crime drama, comedies, foreign films and documentaries), how will this affect my future Netflix recommendations, if at all?

The next Netflix algorithm at work became apparent from the notifications section where Netflix recommended ‘top picks’ for me: Friends, Chef’s Table and Transformers.

 

Netflix also provided me with recommendations based on my past viewing habits, as seen in the following photo:

Netflix’s ‘Because you watched…’

And finally, Netflix also recommended “Movies & TV from the 80s” (as I posted HERE). It seems to know when I was born or (at least) which era I grew up in. This is a collection of Netflix’s recommendations for me: Movies & TV from the 1980s. Admittedly, I was excited to watch Jem and the Holograms again!

Google

After reading Siemens (2013) article, I was enlightened about Google’s ‘knowledge graph’ as an example of “articulating and tracing the connectedness of knowledge” (p.1389). Since I was eating a bowl of raspberries, I tried searching for “amount of calories in 10 raspberries” and came up with the following results on Google’s knowledge graph:

Google’s knowledge graph provided me with a wealth of information on raspberries and cited sources from Wikipedia and the USDA. Although the information is useful and can be obtained without going beyond the initial Google search, why does Google obtain its source information from only Wikipedia and the USDA? Are there other sources that are not listed? Why does the algorithm work this way – what is happening behind the scenes that I’m not seeing?

Since Google is so pervasive and, I dare say, educators and students alike use it to perform a magnitude of daily internet searches, should we question how the information being presented to us is gathered or blindly trust Google as an institution in the search engine field?

Moreover, I checked my topics on my Google+ profile and found it interesting to see the results (as seen in the photo below):

My Google+ topics

As it says, “these topics are derived from your activity on Google sites…” It is interesting to note that I had to manually add in the topics of ‘Anthropology” and “Archaeology” because I have a causal interest in these areas and I felt like Google might cheat me out of potential fascinating web content if I didn’t add them to my list. I also (embarrassingly) felt a little hurt that Google didn’t automatically recognize that anthropology and archaeology are interests of mine. Am I disappointed in the algorithm’s performance? It seems ridiculous for me to have feelings towards this since I’m talking about a machine who is simply “running complex mathematical formulae“, but despite this, I was affected. Is this what Knox (2015) is talking about in reference to the “co-constitutive relations between humans and nonhumans?”


References

Knox, J. 2015. Algorithmic Cultures. Excerpt from Critical Education and Digital Cultures. In Encyclopedia of Educational Philosophy and Theory. M. A. Peters (ed.). DOI 10.1007/978-981-287-532-7_124-1

Siemens, G. (2013) Learning Analytics: the emergence of a discipline. American Behavioral Scientist, 57(10): 1380-1400

Exploring Algorithms, part 1

Exploring Algorithms, part 1

Algorithms on YouTube

I started to explore #algorithms while searching for YouTube videos. In keeping with Christian Sandvig’s Show and Tell, I started to type ‘residential’ and before I could finish, Google Instant came through with this prediction:

Apparently Google thinks I want to see Resident Evil 7 – a horror video game (yikes)… Interestingly, as soon as I typed the ‘i’ in ‘residential’, Google knew that I was searching for info on the residential school system in Canada. Given the last course I took in the MSc programme was digital game-based learning, I’m assuming these predictive results turned to games since I had researched this subject in the past. 

 

 

 

 

 

 

 

 

 

In this POST, I included shots of all the predictive results of my search for residential schools in Canada (or also find it here on Tumblr). As you can see, some of the results branch out to include documentaries about native Americans.


Although I was aware that algorithms were working behind the scenes to tailor personalised recommendations on Facebook, Google, Netflix, Amazon, etc., until this section of EDC, I didn’t realise the extent of the algorithm’s reach and influence. As Knox (2015) points out, Amazon’s algorithm has significant influence over spending habits of consumers as does Facebook ads. In this brief article, Jerry Kaplan discuss the impact of Amazon’s algorithm and the concept of ‘information asymmetry’ where one party has more or better information than the other, creating an invisible imbalance in power. Although they are human-created, these algorithms seem to have the ability to outsmart us and (perhaps) cause us to shell out more cash than we should!

Why did the following ad appear on my Facebook page?

Ad that appeared on my Facebook page.

This ad for Stella Artois appeared on my Facebook page and struck me as somewhat unusual. From my observation, most of the ads that pop up on my page come from (at least I think they do) my recent Google searches and from my personal preferences. Beer, however, and specifically Stella, is rarely something I search for (if ever). Looking closer, I noticed at the bottom it says “Purchase a chalice. Help end the global water crisis.” Is this ad really a call for social action or is it just trying to get me to buy beer?

Upon further investigation, I visited Stella’s website (buy a lady a drink) and discovered their Chalice promotion and partnership with water.org. A ‘limited edition’ Chalice can be purchased for $13.00. Stella’s disclaimer on their website states that $6.25 provides clean drinking water to 1 person for 5 years. Stella Artois will donate to water.org $6.25 for every chalice sold in the U.S. in 2017, up to 200,000 chalices.

Upon reflection, I’ve come to realise that perhaps, in some cases, Facebook ads and the algorithms that create them can be viewed in a positive light in terms of social impact. I taught a marketing class at Durham College this past fall and one of our topics of discussion centred around social responsibility in corporate marketing. Stella’s Chalice programme seems to be participating in this kind of marketing in an effort to aid in the global water crisis. Can algorithms lead individuals and/or organisations to partake in social action for positive change? Again as Knox (2015) points out, we must remember that algorithms are political and biased, leading us to think about “what kind of individuals and societies are advantaged or excluded through algorithms.”

Who benefits?

No doubt that Stella Artois is adding to its bottom line by implementing the Chalice programme, but it seems they are also trying to create a company image of social responsibility. Is Stella’s partnership with water.org creating a positive impact on those who are in desperate need of clean water?

A few years ago, I was involved with an environmental group in my local area: The Enniskillen Environmental Association (EEA). The EEA, who are a handful of concerned citizens, fought Hydro One in order to stop a mega transformer station from being built on the Oak Ridges Moraine – a large water-rich protected area. As this was a David and Goliath type of fight, the EEA didn’t have enough power to combat the enormous wealth and spite of Hydro One and eventually lost the battle. That being said, perhaps my past involvement with water conservation had some influence on why the Stella ad appeared on my Facebook page? Do the algorithmic ‘gods’ somehow know I was involved in social justice practices? I could be reading into this too much, but how far can the reach of algorithms extent?


In preparation for my next cohort of marketing students, I’m thinking about how to incorporate the analysis of algorithms in marketing and the implications of these algorithms in terms of education. What kind of class activities can I create to discover and/or track organisations who participate in social marketing practices and unpack the resulting research in terms of impact for society at large, for the organisation and for education?

Watch my EEA video here:

References

Knox, J. 2015. Algorithmic Cultures. Excerpt from Critical Education and Digital Cultures. In Encyclopedia of Educational Philosophy and Theory. M. A. Peters (ed.). DOI 10.1007/978-981-287-532-7_124-1