WhatsApp importer for NodeXL Pro

Import social media networks from WhatsApp Chats using NodeXL Pro

WhatsApp is a widely used personal messenger application owned by Meta.  WhatsApp allows people to form chats with one or more people.  WhatsApp chats can grow to include large numbers of participants, effectively becoming semi-public group discussions. In some cases, these discussions can be the focus of collective action with civic impact.

WhatsApp is also evolving into the gaps left by other platforms, adding a new “channel” feature that will be friendly for brands and organizations. The WhatsApp business model seems to be in robust shape with a promising future.

In contrast with other forms of social media platforms, WhatsApp chats are more under the control of their authors and members, since people can be selectively invited in and removed by the chat owner. WhatsApp chats may be less discoverable than discussions on other social media platforms that have global search features, but this absence may be a feature as much as a bug, providing a social boundary that is crossed only through invitations from existing members.

As conflict in public social media spaces becomes more common, participation becomes less attractive. In contrast more bounded social media spaces can be more inviting and personal. Many users report that they participate in many WhatsApp chats and that some of these chats are old, involve many people, and have many messages.

NodeXL can now import and analyze the files that WhatsApp exports that contain all the messages in a chat. Version .522 now has an importer that turns WhatsApp chats into social media networks. Check the NodeXL Pro > Data > Import > Import from WhatsApp (Beta) menu.

Creating these files is a simple process.  From any WhatsApp group chat, tap on the chat name and access the Group Info page.

Scroll to the bottom of this page to get to these options:

Select the “Export Chat” option.

This will generate a zip or text file that you can download to your NodeXL PC.

In NodeXL, navigate to the menu NodeXL>Data>Import>From WhatsApp chat file to access this dialog:

Navigate using the browse button to select the chat file you exported from WhatsApp. This file is usually named _chat.txt.

WhatsApp chat files are simple.  Each message is a row in the data set.  Each row has a simple format that is some variation of this:

[date, time] username: message text

In some cases the message text will contain usernames indicated as @username. In WhatsApp, a username is often a telephone number.

In some cases the date and time will have a less common pattern depending on your location and preferences. If your version of WhatsApp chat export does not load properly, please contact us so that we can adapt the importer to your format!

The NodeXLPro WhatsApp importer builds social networks from this data.

WhatsApp users can be implicitly and explicitly connected to one another based on their messages.  

One type of implicit connection is to link each user to the user who authored the prior message.  The sequential position of messages is often but not always an indication of a reply. We call this a “consecutive” relationship or edge type.

Users can also be explicitly connected when they reference the @username of another user.  In WhatsApp an explicit reply is assisted with a drop down menu list of the usernames of the other users in the chat. Some users are represented by user names while others are represented by their telephone number.

We recommend selecting the first two options in the NodeXL Import from WhatsApp dialog.

When you click OK, NodeXL reads and transforms the WhatsApp chat export file into a network.

Users can also be said to link to themselves (called “self loops”) when they post and these links are created optionally when NodeXL imports a WhatsApp chat.

WhatsApp chats can vary widely in size from just two people to several thousand.  Families, friend groups, work teams, and neighborhoods often create WhatsApp chats to share news, requests, and offers. This is a small sample network formed by the NodeXL team to illustrate the ways chat messages can form links to other users.

When these chats contain many messages and many people it can be difficult to review the complete history of a chat to find the key events of peak activity, the most central or influential members, or a summary of the content and links exchanged.

An automated analysis of WhatsApp social networks using NodeXL Pro provides a quick and easy way to summarize and extract the key people, groups, events, and topics discussed in a chat. NodeXL will generate a time series of the activity in the _chat.txt file. It will analyze the messages to build the network of connections based on who posts right after a prior person’s message and who “mentions” other users in the text of their messages. The collection of these connections forms a social media network that often contains structures, sub-groups, and indications of the most influential participants in the discussion based on their location at the center of the network. Here, for example, is a large WhatsApp discussion that took place over a long period of time, with many participants. Clear clusters emerge along with the most central participants.

Looking at WhatsApp chats in this network centered way can enable the study of these computer mediated collective action platforms and the groups that emerge from their use.