How do you make ads less obnoxious? Make them more useful. The more useful they are the more likely someone is to click on them. It’s because of this that Click-Through-Rate (CTR) is the KPI that is measured to optimize ad networks. The author of this paper have a proposal of how to predict and therefore optimize CTR.
Effendi, M. J., & Ali, S. A. (2017). Click Through Rate Prediction for Contextual Advertisment Using Linear Regression. arXiv preprint arXiv:1701.08744.
Continue reading “CTR for contextual w. Linear Regression – Effendi”
Fuck me Google Analytics (GA) has a lot of information in it. How do we make use of it meaningfully? In this paper, two researchers propose a system for us to bridle the stallion of GA and turn it into a driving force for change in the world of marketing.
Chaffey, Dave, and Mark Patron. “From web analytics to digital marketing optimization: Increasing the commercial value of digital analytics.” Journal of Direct, Data and Digital Marketing Practice 14, no. 1 (2012): 30-45.
Continue reading “Digital Marketing Optimisation: Increasing value of digital analytics – Shaffey & Patron”
Trawling through UN debate transcripts and comparing them against the countries voting patterns? Yeah, that does sound pretty terrible. Well Gurciullo and Mikhaylov are using Doc2Vec to make it easier!
Gurciullo and Mikhaylov. “Detecting policy preferences and dynamics in the un general debate with neural word embeddings.” arXiv preprint arXiv:1707.03490(2017).
Continue reading “Detecting policy preferences and dynamics in UN general debate – Gurciullo & Mikhaylov”
So you’ve mucked about with Gensim and found that Word2Vec is incredible and you want to use it for everything. You google ‘best ways to cluster word2vec’ and you find like… two githubs [here and here] that don’t really have great explanations.
I was like you, so I did some tests! Github Here!
TL,DR: Use SKLearn’s SpectralClustering on the vector dataset.
Continue reading “Best Ways To Cluster Word2Vec”
In writing my paper, I aimed to argue that ML models could replicate the open and closed coding of two researchers. Building better mousetraps by Lin aims to argue that the more general objections that researchers have raised against using bigger data in sociological research are false. Lins a smart guy and while not proposing any brand new arguments, makes a great summerisation of the arguments out there.
Lin, Jimmy. “On building better mousetraps and understanding the human condition: Reflections on big data in the social sciences.” The ANNALS of the American Academy of Political and Social Science 659, no. 1 (2015): 33-47.
Continue reading “On Building Better Mousetraps and Understanding the Human Condition – Lin”
Like Frank Herbert’s Dune? Yeah me too. Want to understand more about the political machinations and structure building with a little bit of that sweet sweet time hijinks? Well you came to the right place. Rudd questions whether the history that Muad’Dib creates and builds his empire within if linier and how we can best understand the way he adapts his predesessing emporor’s empire.
Fun the whole race-concious family!!
Rudd, Amanda M. “Paul’s Empire: Imperialism and Assemblage Theory in Frank Herbert’s Dune.” MOSF Journal of Science Fiction 1, no. 1 (2016).
Continue reading “Paul’s Empire: Imperialism and Assemblage Theory in Frank Herbert’s Dune – Rudd”
You know that feeling when you’ve got a kickass linguistics model but just showing the text outputs to your boss/partner/mother just isnt getting the reaction that you wanted? I do. I’ve just finished putting together a manuscript using data data models just like this to submit to journals and wanted to share some of the visualisations with you in case you were at a loss like I was.
Continue reading “Data Visualization for Gensim LDA and Word2Vec”
Regardless of whether you code in R or not, this is a great paper. It breaks down the processes and ideas that you’ll use in topic modeling and data pre-processing into clear steps with just enough maths that you’ll be wishing you took it for a-level!
Great paper, read it.
Karl, A., Wisnowski, J., & Rushing, W. H. (2015). A practical guide to text mining with topic extraction. Wiley Interdisciplinary Reviews: Computational Statistics, 7(5), 326-340.
Continue reading “A practical guide to text mining with topic extraction – Karl, Wisnowski, & Rushing”
Always wanted to bedazzle your friends with personality insights from their writing but didn’t have the money for a Watson api? Were you like me and hoping for some insights on how to use corpora for sociological research? Well buckle up, there’s a paper here for you!
Argamon, S., Koppel, M., Pennebaker, J. W., & Schler, J. (2009). Automatically profiling the author of an anonymous text. Communications of the ACM, 52(2), 119-123.
Continue reading “Automatically profiling the author of an anonymous text – Argamon, Shlomo, et al.”
Following on from last week’s methodology, the time has come to talk about findings! I’m really excited to talk about the findings here as they are attempting to bucket the methods and strategies that Men’s Rights Activists are using online both to create their own identities and to convince others by progressing their arguments. Their research questions were;
Lets dive back in!
Schmitz, R. M., & Kazyak, E. (2016). Masculinities in Cyberspace: An Analysis of Portrayals of Manhood in Men’s Rights Activist Websites. Social Sciences, 5(2), 18.
Continue reading “Masculinities in Cyberspace – Schmitz & Kazyak – Part 2 (Methodology)”