BotGBIF: A tool to query GBIF data in natural language

Abstract

BotGBIF es un prototipo de chatbot que facilita el acceso a datos de biodiversidad del Sistema Global de Información sobre Biodiversidad (GBIF) mediante consultas en lenguaje natural. Utiliza Streamlit para la interfaz, la API REST de GBIF para obtener datos y GPT-4 como modelo de lenguaje. Su objetivo es simplificar la interacción con los datos para usuarios sin conocimientos técnicos, permitiendo consultas intuitivas y accesibles para investigadores, conservacionistas y el público general.

This project introduces BotGBIF, a prototype chatbot designed to enhance user access to biodiversity data from the Global Biodiversity Information Facility (GBIF) through natural language queries. Utilizing the Python Streamlit library for the frontend, the GBIF REST API for data retrieval, and the capabilities of GPT-4 as a large language model (LLM), BotGBIF aims to simplify the interaction with complex datasets for users who may lack technical expertise in utilizing the GBIF portal. By integrating the GBIF API with GPT-4, the chatbot allows users to ask questions in everyday language and receive intuitive responses, thereby making biodiversity data more accessible to a range of stakeholders, including researchers, conservationists, and the general public.