Text Filter for DataFrame or Multi Select with ML generated Tags?

My app is going to be used for analytics derived from Android LogCat Logs. Part of the app actually has a lot to do with mining the logs as well. I provide a dataframe at the top of the app and some more basic filters to narrow down to the log lines that they need for analysis. I am struggling with how I approach providing a filter for the ‘log message’ column. The ‘log message’ is the text written for each process id and thread id.

  • Does Streamlit have a text filter?
  • I know there is a text_input, but I am not sure if that can be used to kick off a query that could filter the dataframe?
  • Would I be better off using an unsupervised ml model to generate tags and provide those in a multi-select that then filters the dataframe?

Looking for some ideas and suggestions from the community.