FAQ
Answers to questions the SMM team has received from users. If your question isn't here, email smm@lists.illinois.edu.
The hosted SMILE Playground and NCSA deployments are offline (as of October 2026). The answers below apply to any SMILE instance, including one you run yourself.
Watch the SMILE overview video. Then, since there is no hosted instance, start your own with Run SMILE with Docker Compose and try things out. Our publications describe the design and intended use of the tools in more depth; see Cite our work.
When you run SMILE yourself, all collected data and results stay in your own deployment's storage (the MinIO volume on your machine or cluster), under your control. Each SMILE instance shows its privacy policy at /privacy-policy. SMILE is also governed by the University of Illinois Terms of Use, and it must not be used commercially without first speaking with the university.
Data you collect from social media platforms remains subject to those platforms' terms of service. See Why does SMILE ask me to authorize my Twitter or Reddit account?
SMILE assumes you already have an account on the platform you want to collect data from. If you don't, sign up at x.com or reddit.com first.
When you click Authorize in SMILE, you are sent to the platform's own website to log in and approve SMILE. You never give SMILE your username or password. Instead, the platform confirms who you are and gives SMILE a token that lets it collect data on your behalf. This is the standard way third-party apps connect to social media platforms (see Twitter's explanation of connected apps).
Because the requests are made as you, you are bound by the platform's terms of service and are responsible for any violation of them. You can revoke SMILE's access at any time from your account settings on the platform.
Twitter (now X) has heavily restricted API access since 2023. Whether Twitter collection works depends on the API access level of the developer account behind the SMILE deployment you are using.
The full message looks like:
{"ERROR":{"name":"FetchError","message":"request to http://localhost:5050/graphql failed, reason: connect ECONNREFUSED 127.0.0.1:5050", ...}}
SMILE couldn't reach its data server, the component that talks to the social media platform APIs.
If the deployment is otherwise working, this usually means the connection was unstable or the request was very large. End your session, start a new one, and retry with the default settings (don't open Advanced). You won't lose existing data. If it keeps happening, email us with:
- the keyword you searched,
- the source (tweets, Twitter timeline, Reddit posts, or Reddit comments),
- how many posts you requested,
- whether you used the default or advanced settings,
- whether you saw any preview.
On your own Docker deployment, check SMILE_GRAPHQL_URL. It must use an address the SMILE container can reach, such as your machine's IP address. localhost won't work because, inside a container, it refers to the container itself. See Run SMILE with Docker Compose.
This happens when you imported a CSV file without telling SMILE which column contains the text to analyze. When you import the file, choose a category of Others..., tick the column(s) you want to analyze under Which column(s), pick the analyses, and click Set Headers:
If the column is configured correctly, selecting the file in any analysis shows a preview with only those columns, and the analysis form includes a Select Column to Analyze option:
The data is fine; Excel is opening it with the wrong text encoding. To import it correctly:
- Open a blank Excel workbook.
- Go to Data → From Text (or Get Data → From Text/CSV in newer versions).
- Set the file origin to Unicode (UTF-8).
On macOS you can also open the CSV in a text editor, which displays the characters correctly.
Too many requests have been made to Twitter from your account. Wait 15 minutes and try again, and make sure you aren't pulling Twitter data from any other application at the same time.