Analysing Twitter Users' Timelines

These instructions describe how to download users' timelines (i.e., previous tweets) with Webometric Analyst and some simple analysis methods.

1) Download Webometric Analyst to your PC using the link at the top of this page, if you haven't already done so.

2) Create a plain text file using Windows Notepad with a list of the Twitter usernames that you would like to analyse. All (as far as Twitter allows) of the tweets of these people will be downloaded, including their retweets. Here is an example file of tweeters.

3) Start Webometric Analyst, click Classic Interface in the Wizard and then click the Twitter tab in the main interface (see the arrow below).

Twitter section of Webometric Analyst

4) Click on the button at the bottom of the image below and select your file of usernames. Follow the instructions to get permission to download tweets and wait for the program to finish. If you have a lot of acounts or very active acounts then it may take several hours.

5) When the program finishes, the tweets will be saved in a large text file. You can load this into a spreasheet like Excel to view it or can process it with other Webometric Analyst options.

6) To get a random selection of tweets (for a content analysis) and basic statistics about the tweets (example), click the Count Tweets per Query... button and select the file of tweets.

7) To analyse word frequencies in your tweets, download and install Mozdeh and then click the Convert Twitter Files To Mozdeh Format to export the timeline tweets to Mozdeh. Note that when you convert to Mozdeh in this way the following happen

Important: Check the "Filter out duplicate tweets with identical text AND from the same tweeter" option because the same tweet may occur in different timelines.

Note that if you have a collection of related Twitter accounts then you may get retweets from one of the accounts appearing in other accounts.

Important: To EXCLUDE all retweets from your analysis in Mozdeh then (1) list the usernames that you searched for in the Twitter Names of Tweets to extract dialog box, separating each username by a | character (not case sensitive). See below for an example and the example lists in this file. Also (2) check the "Filter out duplicate tweets with identical text AND from the same tweeter" option because the same tweet may occur in different timelines. (Using this in conjunction with the Ignore Retweets option can cause lost tweets).

Tweet extraction dialog box

See the instructions on the Mozdeh site for ways in which the tweets can be analysed.