Ad verba per numeros

Tuesday, May 1, 2012, 06:21 AM
If you are a regular reader of this blog you probably know my position regarding predicting elections with Twitter. In case you don't know it: you cannot predict elections! [1] [2] [3].

The thing is that time and again I read about the endless possibilities of Twitter regarding politics and, specially, electoral predictions.

Time and again I see the same papers cited; papers that, needless to say, claim that predictions are easy and absurdly accurate.

And time and again I see new papers stubbornly ignoring the (few) negative results showing that predictions are neither that easy nor possible with the current "state of the art" methods.

Therefore, I've reached my limits; I've felt that "someone's wrong on the Internet" moment and I've written a short survey on the topic.

More than a survey the paper is an annotated bibliography covering anything written on the topic of predicting elections with Twitter. Both positive and negative results are commented but more importantly I suggest what I see as key future lines of research.

The survey is available as a preprint at arXiv:

"I Wanted to Predict Elections with Twitter and all I got was this Lousy Paper" -- A Balanced Survey on Election Prediction using Twitter Data

Predicting X from Twitter is a popular fad within the Twitter research subculture. It seems both appealing and relatively easy. Among such kind of studies, electoral prediction is maybe the most attractive, and at this moment there is a growing body of literature on such a topic.
This is not only an interesting research problem but, above all, it is extremely difficult. However, most of the authors seem to be more interested in claiming positive results than in providing sound and reproducible methods.
It is also especially worrisome that many recent papers seem to only acknowledge those studies supporting the idea of Twitter predicting elections, instead of conducting a balanced literature review showing both sides of the matter.
After reading many of such papers I have decided to write such a survey myself. Hence, in this paper, every study relevant to the matter of electoral prediction using social media is commented.
From this review it can be concluded that the predictive power of Twitter regarding elections has been greatly exaggerated, and that hard research problems still lie ahead.

The first part of the title is deliberately provoking but I feel it's moment for the research community involved in this area to think twice before making flamboyant claims on print.

Needless to say, I'll be happy of hearing your comments and you can find me on Twitter: @PFCdgayo.

Back Next