Enabling Public Discourse Research with Telegram Data – No Coding Required

By Philip Mai

We are thrilled to announce the release of the new Telegram data collector in Communalytic, our newest computational social science research tool. The new data collector will provide academic researchers with a systematic way to access and study public discourse on one of the fastest growing social media apps.

If you are not familiar with Telegram, it is an encrypted app similar to Meta’s WhatsApp and Facebook Messenger. Launched in 2013 by entrepreneurs Nikolai and Pavel Durov (also founders of Russian social media platform VK), Telegram is now one of the top-5 downloaded apps worldwide with over 700 million monthly active users.

But Telegram is more than a messenger app. It also hosts numerous groups and channels that allow for one-to-many and many-to-many types of communication. These public groups and channels attract millions of users from all walks of life, including governments, politicians, pundits, news organizations, activists and many others. For example, when Russia invaded Ukraine earlier this year, Telegram “became the go-to app for Ukrainians” and while at the same time it also emerged as one of the “main vectors for invasion disinformation”. Telegram’s ease of mixing the public and private, and its lack of a formal mechanism to report illegal content have attracted terrorists and fines from governments over the years. 

Because of this presence of diverse communities and controversial issues, Telegram is beginning to attract the interest of researchers who are keen on conducting independent research in the public interest in areas such as mis/dis-information, hate speech, online extremism to name just a few. However, up until now, only researchers with programming skills have been able to access publicly available data from Telegram. With the introduction of the new Telegram collector feature in Communalytic, we hope to address this access gap by providing an easy-to-use, web-based interface for data collection and analysis of Telegram data.

If you are interested in learning more, here are a few helpful links to get you started with Telegram research via Communalytic:

——————— 

About Communalytic

Communalytic is a computational social science research tool for studying online communities and discourse. It can collect, analyze, and visualize publicly available data from various social media platforms including Reddit, Telegram, Twitter, and Facebook/Instagram (via CrowdTangle), or from your own CSV or JSON files.  

A suite of data analytics modules designed for research

Communalytic contains a suite of data analytics modules including: 1) a Toxicity Analyzer via Google Perspective API, 2) a Sentiment Analyzer via libraries such as VADER (EN), TextBlob (EN, FR, DE) and Dostoevsky (RU), 3) a Bot Analyzer via Botometer API, and 4) a Network Analyzer. These modules can be used to: identify and examine anti-social interactions, assess sentiments in online discourse, detect Twitter bots, identify influencers, map shared interests among online actors by examining what topics or links they shared, study the spread of mis- and dis-information as well as look for signs of possible coordination among seemingly disparate actors. 

One of the unique features of Communalytic is the ability to generate and visualize so-called “signed” networks via the built-in Network Analyzer module. A signed network is a network with edges that contains additional information such as positive or negative signs or scores (weights). Communalytic builds a signed network by assigning toxicity scores and/or sentiment polarity scores as weights to edges in the network. This feature can be used to identify and visually highlight interactions of interest (e.g., anti-social interactions) within the network so that they may be examined in more detail.

In addition, if users are working with Twitter data, they also have the option of running the Bot Analyzer and adding a bot probability score as an attribute to the nodes in the network generated by Communalytic. This feature can be used to identify and visually highlight interactions of interest (e.g., Twitter accounts that might be bots) within the network so that they may be examined in more detail. (For more details see: the Commnalytic FAQ and Tutorial page.)