Open Framing

A tool for analyzing multilingual media frames using AI technologies

Open Framing AI (OFAI) conducts systematic framing analysis of multilingual texts. The tool draws from social and computer science theories and methods.

If your goal is to make sense of large, textual datasets, our AI-assisted tool will quickly help identify frames that dominate the discussion. Check out the AI & Emerging Media (AIEM) research team behind the tool.

If you use our tool, please remember to use our reference below. Thank you!

Guo, L., Su, C., Paik, S., Bhatia, V., Akavoor, V. P., Gao, G., ... & Wijaya, D. (2022). Proposing an Open-Sourced Tool for Computational Framing Analysis of Multilingual Data. Digital Journalism.

How OFAI works

How to get started

Conduct a five-step framing analysis
Find and collect data about a subject matter you're interested in analyzing frames. Common datasets come from news articles, social media posts, or advertisements.

Get started

If you're not sure...
Don't fret! We got you covered. You can use the OFAI tool in many ways and do all sorts of analysis beyond framing. See our step-by-step guide here.

Guide

Also, check out our demo page which has case studies of the data output and visualization capabilities.

Demo

What is Framing?

Framing refers to the selection of some aspects of a perceived reality and make them more prominent in a communicating text such as a news article and a Tweet.

Examples

When some news media emphasize the mental illness of gun shooters over other other aspects of gun violence in covering the issue, this is framing.

When you choose to purchase a yogurt product that is advertised "90 percent less fat" rather than the one saying "10 percent fat," this is when the framing effect occurs.

OFAI is advantageous because:

1. Best of two worlds
You will access the state-of-the-art computer science methods, and get results that are valid and interpretable according to communication research standard.

2. Transparent
All the algorithms used in the analysis are open to users. There is no black box problem. See our source code here.

3. User-friendly
The tool is designed to be click-and-play, so users with relatively little programming knowledge can still run complex algorithms such as deep learning.

4. Multilingual
The tool can be applied to analyze texts of different languages in addition to English.