Christmas on Twitter

Return to the visualization »


Thanks for visiting the Xmas Tweets 2012 project! People regularly use Twitter to report what they are thinking, doing, or planning. So, not surprisingly, they use it to talk about what they want for Christmas, too! On this website, we've mined over 1.3 million wishes that people have posted on Twitter and collected stats on some of the most popular and interesting ones.

Why study Christmas wishes?

What people wish for can tell us a lot about what they value. As a result, wishes can provide a way of assessing the items, products, and even intangibles (e.g., "jobs", "family", "vacation") that are particularly important to people in a given year.

Wishes are also, not surprisingly, an indicator of what people will buy. This is more true for some things (e.g., "Xbox") than others (e.g., "alien"), so we have to be careful when we apply this principle. Nonetheless, we've found that there is a strong correspondence between how many people wish for certain items and how many of those items are sold each Christmas.

Thus, overall, by studying Christmas wishes we hope to better characterize and understand ideas that have a pronounced presence in the social consciousness and also to explore a new mechanism for predicting consumer buying behavior.

From tweets to wishes

Going from raw tweets to proper wishes isn't a trivial process. The first step is to run a program that reads tweets out of Twitter (via the "firehose", a real-time feed of tweets) flag tweets that contain words indicative of Christmas wishes: "want", "Christmas", and "wish", for example. Then a computer program with a more precise filter looks at these flagged tweets and extracts individual wishes. These are then examined (by another program) to find the most popular wishes people have made.

In order to determine gender, we used census data which tells us how likely a name is to be associated with a particular gender. "Wendy", for example, is 100% associated with females - no males are reported as having that name. Since each wish is linked to a person's Twitter account, if that person provides a name, we can determine the probability that that wish was made by a male or a female.

The Xmas Tweets team

This website represents ideas contributed by everyone in the Network Dynamics Lab. Several members deserve special mention for their work on specific components:

If you have feedback for us on the site or ideas that we should explore, we'd love to hear your thoughts. Please drop us an email.

❆ Happy holidays!! ❆