Twitter Tips – Dr. Seuss Style

The folks over at HootSuite (my current favorite Social Media Management Dashboard) recently came up with this wonderful Dr. Seuss-Inspired Guide to Twitter. Very nice, humorous way to showcase several Twitter Tips for brands (remember #hkuiom – you are also a brand!).

 

What makes a tweet go viral?

There is a lot of recent research studies that investigate what factors influence the popularity of memes on social networks. Much of this research analyzes twitter posts and has identified many reasons why certain tweets go viral. These include factors related to the tweet itself (e.g. how controversial the tweet is) and factors related to the tweeters (e.g. number of followers, influence, and frequency of posting). New work says that ‘going viral’ is a random process.

Visualizations of meme diffusion networks for different topics.

This new study uses an agent-based model to study this phenomenon. This model simulates message sharing on a social network and incorporates two key characteristics of such a context: users have limited attention spans and can only view a portion of all tweets.

The predictions of our model are consistent with empirical data from Twitter, a popular microblogging platform. Surprisingly, we can explain the massive heterogeneity in the popularity and persistence of memes as deriving from a combination of the competition for our limited attention and the structure of the social network, without the need to assume different intrinsic values among ideas.

Or in other words, the pattern of twitter memes can be replicated in the absence of tweet or tweeter based factors.  This raises interesting questions regarding the direction of causality – do tweets go viral because of certain factors, or is the popularity of posts on social networks a random process and we find mere correlations in our bid to find explanations. While some say that this correlation versus causation conundrum can be solved only empirically, others say that a controlled, experimental approach is the way to go.

Suppose, Menczer says, that in his study he randomly assigned different colours to each tweet. If red tweets ended up being the most popular, one could argue that colour was a predictive factor for success when in reality the popularity of red-coloured tweets was coincidental.

Read more at  Going viral on Twitter is a random act – tech – 13 April 2012 – New Scientist and Competition among memes in a world with limited attention : Scientific Reports : Nature Publishing Group.

References:

Weng, L., Flammini, A., Vespignani, A. & Menczer, F., Competition among memes in a world with limited attention, Scientific Reports 2, Article number: 335 doi:10.1038/srep00335

Stock market prediction using Twitter

A recent paper presented at the ACM International Conference on Web Search & Data Mining uses a simulation to show that a trading strategy based on Twitter conversations can outperform the Dow Jones and basic trading strategies. Using twitter and stock market data over a four-month period, the study also finds that the number of distinct twitter conversations about a stock is a strong predictor of stock trading volume. A higher number of distinct conversations is also positively related to higher stock prices while lower number of conversations can predict a lower stock price.

 

The main trading room of the Tokyo Stock Excha...

Image via Wikipedia

 

 

 

While past research has looked the sentiment, positive or negative, of tweets to predict stock price, little research has focused on the volume of tweets and the ways that tweets are linked to other tweets, topics or users. Further, past work has mostly studied the overall stock market indexes, and not individual stocks.

 

A key limitation of the study is that the market lost value during the examined time period. Thus the Twitter based strategy outperformed other strategies by losing the least amount.

 

For the study, the researchers simulated a series of investments between March 1, 2010 and June 30, 2010 and analyzed performance using several investment strategies.

 

Read more here and here. See the paper here.

 

 

 

Twitter fuels a bank run

Swedish banks operating in Latvia were recently victim to a social media-fueled bank run. Analysis by Orgnet.com shows the social graph of Twitter users who tweeted and retweeted rumors about the bank. The image highlights the Swedbank‘s central role in tweeting denials and trying to control the spread of rumors. One can clearly see two sub-networks at play – the first being a highly connected network of rumor-mongers, the second a highly centralized network of denials.

 

bankrunupdate.png

Image via ReadWriteWeb

 

Download the original pdf here: Twitter Bank Run.

Read more at Did A Twitter-Fueled Latvian Bank Run Start With One Account? [UPDATED].

When stuff tweets

Twitter will soon be awash with tweets by stuff (inanimate objects). Twine, a $99 wifi connected box, uses inputs from environmental sensors and a set of action rules to tweet different messages. Users can set up the action rules online. Thus the day is not far when you will receive a tweet from the toaster that the toast is ready! The box has various inbuilt sensors and allows endless extensions.

The battery-powered box contains sensors for temperature and vibration, a magnetic switch and a moisture sensor. Pretty much anything else can also be added to the contraption. One backer plans to outfit Twine with weight sensors and use it to notify him when the ice machines he operates need refills. Another will use a magnetic door sensor to receive a message when UPS stops by. Others say they will keep track of their pets, heating systems and garage doors using the device.

 

Image via Mashable

 

Read more at Twine: The Revolutionary Box That Can Make Your Appliances Tweet.