Kindness List

Kindness Story List

Generative AI can collect kindness stories from media and help us learn from kind people.

Index ID Source Year Category Story Summary
Index ID Source Year Category Story Summary

Search a keyword to shorten the story list.

Procedure of the kindness database generation Actual Deck

The kindness story is collected by the AI tool Gemini after making a request prompt to list the kindness news from a news media and limit the length of the content and the range of the event dates. The list is accumulated in an excel file. The excel file is further converted into a json list. The Python pandas dataframe is used to organize the information in a json list by adding a unique ID for each story and converting the information into a dictionary with keys of Source, Kindness Category, and Story content.

The Large Language Model checkpoint library "facebook/bart-large-xsum" is used to create a summarizer pipeline. By feeding the story dictionary into the summarizer, an additional category of "Summary of each story" is produced.

The jQuery DataTable is used to display the list of kindness stories. A click button provides a user interface to access the content of the selected kindness story in a separated webpage.

The KerasCV Stable Diffusion is used to generate AI images using text prompt.