Dr. Pinnaree Tea-makorn, in a collaborative research with Punyanuch Maitreenukul, Assistant Professor Stefano Starita, and Assistant Professor Pavitra Jindahra, delivered a presentation on “Exploring Data Analytics Climate Startups Using BERTopic.” The data and findings on this research can be used by individuals who have access to the paper, providing them with summary statistics for companies specializing in climate tech startups that leverage data analytics. Applying data analytics with Net Zero Insights, Dr. Pinnaree shed light on the transformative potential of AI and data analytics in addressing climate change. Net Zero Insights “operates a global database that tracks startups from pre-seed to exit and SMEs developing innovative products, services, or technologies explicitly contributing to at least one of the six environmental objectives of the EU taxonomy for sustainable activities,” according to LinkedIn. Dr. Pinnaree addressed the multifaceted ways in which AI and data analytics can contribute to climate change mitigation, encompassing the optimization of low-carbon technology and the accurate prediction of weather forecasts. For example, Google’s Deepmind’s AI model, “Graphcast,” was trained on around 40 years of historical data to make an accurate weather forecast around the world. While there was a hiatus in investing in clean tech for a decade, marked by minimal or very minor investments, Dr. Pinnaree said that investment in the clean tech sector resumed between 2018 and 2020. During this period, the percentage of investment in climate tech, relative to all startup investments, increased. This resurgence in interest coincided with a rebranding, with the sector being referred to as climate technology. Venture capital investments in climate tech and AI/data analytics startups have attracted considerable interest, with a staggering $92 billion invested in 2022, complemented by an additional $120 billion directed towards AI and data analytics during the same period. Dr. Pinnaree emphasized that investments in AI and data analytics for climate change cover the following areas:

1. Research and Development of low-carbon technologies, such as photovoltaics and batteries.
2. Planning and design of systems, including urban infrastructure, carbon markets, and smart urban planning.
3. Enhancement of operational efficiency and optimization solutions, addressing challenges in electricity grids, energy efficiency, and weather forecasting.
4. Designing policies, monitoring, and enforcing measures, including tracking greenhouse gas emissions and infrastructure mapping.

Dr. Pinnaree provided valuable insights into the data and funding rounds from the Net Zero database, encompassing data from 55,000 organizations across various continents; 5964 companies were used for analysis. The application of BERTopic for topic modeling in patent analysis emerges as a significant aspect of their research. She further explained the effectiveness of the BERTopic algorithm which applies a streamlined approach to topic modeling, combining multiple steps into one, thus enhancing usability and efficiency. She elaborated on the utilization of embeddings to convert words or sentences into numerical data for computer analysis, including the mention of PatentSBERTa, a specialized embedding for innovation analysis. She explained that embeddings are a way of converting words or sentences into numbers that computers can understand, allowing for analysis and insights. The analysis extends to identifying sectors and funding trends within data analytics startups. Dr. Pinnaree shared the findings of the analysis, revealing the identification of 38 sectors. Notably, the predominant sectors in data analytics startups are agriculture, supply material sustainability, and energy and power monitoring, respectively. Other findings include the prevalence of startups in California, particularly concentrated in Silicon Valley. Dr. Pinnaree delved into Sustainable Development Goals (SDGs) and funding details, with a majority of funding rounds occurring in the seed stage and total funding ranging from $1-10 million. During the Q&A session, audience questions shed light on the potential of ChatGPT for climate data analysis. While ChatGPT hasn’t been integrated into their current research, Dr. Pinnaree delved into the prospect of incorporating ChatGPT in the future. The focus would be on utilizing ChatGPT for tasks such as topic modeling and labeling in the analysis of climate data. In conclusion, Dr. Pinnaree Tea-makorn, along with her esteemed collaborators, has presented a comprehensive exploration of data analytics in climate startups using BERTopic. The wealth of data and findings shared in this research holds significant value for those who have access to the paper, offering essential summary statistics for climate tech startups utilizing data analytics. Investors, analysts, and entrepreneurs can also leverage these methods and findings to analyze extensive datasets, contributing to informed decision-making in investments and the formulation of climate-related policies.