Single Text Classifier Demo

In this section you test the relevance classifier with written texts.

Usage:

- Type a tweet-like text in the textbox.

- Then press Enter.

Select classification model

Prediction threshold

This value sets a threshold by which texts classified flood or fire are accepted, higher values makes the classifier stricter (CAUTION: A value of 1 will set all predictions as none)

0 1

Text

Enter tweet-like text

Examples

Classification

Classification Score

Relevance Classifier Demo

This is a demo created to explore floods and wildfire classification in social media posts.

Upload .tsv or .csv data file (must contain a text column with social media posts), next enter the name of the text column, choose classifier model, and click 'start prediction'.

Upload CSV or TSV File

Text field name

Select classification model

Prediction threshold

This value sets a threshold by which texts classified flood or fire are accepted, higher values makes the classifier stricter (CAUTION: A value of 1 will set all predictions as none)

0 1

Select an ID cell in dataframe to view Embedded tweet

Textbox

Dropdown

	Texts	event_label	model_score	IDs

Question Answering Demo

This section uses RAG to answer questions about the relevant social media posts identified by the relevance classifier

Usage:

- Select queries from predefined

- Parameters for QA can be editted in sidebar

Note: QA process is disabled untill after the relevance classification is done

QA model

Aggregation method

Batch size

Choose between 50 and 500

50 500

Number of results to retrieve

Choose between 1 and 10

1 10

Select at least one query using the checkboxes

Add custom queries like the one above, one at a time

Summary

Textbox

1	2	3

Built with Gradio logo