Hey! let's play something fun with the twitter archive of all my tweets during 2015.

Twitter is one of my social media I have been actively using. Since they allow us to request and download our own twitter archive which includes all of your tweets since your very first tweet till the latest one, so I came up with an idea that it could be geekily fun if I did something further with these data. Then I sat down and wrote the code, and finally have made some data visualizations from the twitter archive in various and little interactive ways.

Basically, data visualization obviously is a way that we present data in a form that is more human-readable, more informative and more expressive. It makes the data easily and quickly understandable and could even lead us to discover some hidden information we have never seen before.

Time Distributions

The number of tweets versus time distribution in various dimensions are shown below. They could exhibit the usual behavior of my use of twitter or even daily life during 2015.

Time of Day

Day of Week




The node-link graph below depicts the users who has been mentioned in my tweets during 2015 and the number of tweets which associates to each of them. Nodes represent mentioned users, whereas links represent participation/co-occurrence in tweets. The weight of a link simply illustrates the number of tweets involved by two users on both ends of the link. The wider link, the more number of mentions. And the length of a link represents the same thing but in a reversed way.

All the nodes can be moved arbitrarily by dragging them gently. Sometimes, the nodes are not properly spread out at the beginning. You may give it a little shake then it will get better. And you may double click on a node to make it unmovable.

Unlike an adjacency matrix, this will be instantly understood without a need of scratch paper or imagination in your head. Regarding this graph, is it possible that friend clusters might have been easily revealed?

Retweets & Domains tweeted

This bubble chart encodes the number of tweets in the area of circles; the bigger the circle, the larger the number of tweets. Although harder to read the value from the size of a bubble, it can pack hundreds of values in a limited space. Each bubble represents a username whose tweets were retweeted or a domain related to links I tweeted during 2015. The former are blue and the latter are red. The size of bubbles vary according to their number of related tweets/retweets.


  • Data processing is done at the client side (in web browser).
  • Timezone 'America/Los_Angeles' was used to process the data.