[Workshop] Next-Gen Cleantech Solutions

Mining Insights from Media and Patent Data with Natural Language Processing (NLP) and Large Language Models (LLMs)

June 7, 2024 - We have just published the presentation and tutorial notebook preview! Check out our tutorial 1 & 2!

May 26, 2024 - Media and patent datasets can be downloaded from Kaggle.

Image by GPT-4o

Introduction

At a time when tackling environmental challenges is of paramount importance, the cleantech industry plays a central role in promoting sustainable solutions. However, technological innovation in the cleantech sector requires a deep understanding not only of the technologies, but also of the market requirements. This information is usually embedded in a large amount of patent and media data, which is difficult to analyze manually to effectively capture the development trend. Using Natural Language Processing (NLP) and the latest advancements in Large Language Models (LLMs) is a natural choice to accelerate innovation. In this workshop, we will share our insights gained in solving this task. Several presentations on various relevant topics will be offered, followed by a hands-on session where participants can try out our LLMs-powered cleantech question-answering and recommendation system.

Objectives

Promote the next generation of NLP enthusiasts and innovators in the cleantech industry
Show how NLP and LLMs can be used to accelerate innovation with media and patent text data

Part 1: Decoding Cleantech

Expert techniques for analyzing patent and media data including NLP for innovation intelligence, LLMs and their application in the cleantech industry, RAG and its application for cleantech innovation, and LLMs-augmented recommender systems for cleantech innovation. (Image by GPT-4o)

Part 2: Emerging Visions in Cleantech

Insights from patent and media data (Image by GPT-4o)

Part 3: NLP in Action

Practical coding for patent and media data analysis in cleantech (Image by GPT-4o)

Acknowledgement

We thank Prof. Dr. Peter Egger (ETH Zurich) for sponsoring this event and Prakhar Bhandari (ETH Zurich) for preparing the patent dataset.

Page updated

Google Sites

Report abuse