According to the UN between 720 million and 811 million people in the world faced hunger in 2020, around one in 10. This was exacerbated by the pandemic and the Ukraine invasion. To mitigate the consequences there is need for farmers across the world to build resilient systems and practice modern farming methods.

With this in mind we decided to build a search engine for Agricultural science journals to help farmers access research easily.

What it does

It is a vector-based search engine for agricultural science journals. It uses pytorch transformers and faiss indexing to giving the most relevant results based on the search query. It also uses Google translation to detect language of query and to translate results to that language from the default english.

How we built it

Using streamlit for the web interface and Pytorch for the model.

Challenges we ran into

getting the translation library to work

Accomplishments that we're proud of

Collecting the journals and building a useful search app.

What we learned

pygoogletrans and deep translator

What's next for Agro Science Search

Adding more agricultural data to help farmers stay up to date with relevant information.

Built With

Share this project: