Analyzing Citizen’s Sentiment on the Energy Transition using Natural Language Processing

A sentiment analysis on the energy transition, commissioned by the World Energy Council and carried out by Omdena.

The world is in the midst of an energy transition. This massive shift aims to move away from reliance on fuels that are destructive to the climate, the environment, and people’s well-being. The goal established by the UN is to “ensure access to affordable, reliable, sustainable and modern energy for all” by 2030. While governments, energy companies, and activists dominate the headlines, the progress with infrastructure and technology won’t be sufficient. A successful energy transition for the good of all humanity depends on the action of individuals. Together with the World Energy Council, the world’s leading member-based global energy network, Omdena explored the use of AI in understanding how people around the world perceive this energy transition and their role in it.

“If the world is truly to achieve a state of zero-carbon energy within the next three decades, then it is realistic to say that the role of each individual person and citizen in each community on the surface of the planet forms an integral part of this journey.“ — Sue Stevenson, Director of Strategic Partnership and International Development, Barefoot College

How each one of us views the steps in this energy transition likely depends on our personal perspective. For instance, while an increase in home fuel bills might be a mere inconvenience for an affluent family, it might push someone in a marginalized community into poverty. A gasoline tax to subsidize renewable energy efforts will be applauded by some and protested by others, as was the case with the “yellow vest movement” that ignited in France in 2018. Though outlawing the use of fossil-fuel-based generators will be irrelevant for someone with reliable access to electricity, it may cast those living in energy poverty into darkness.

Knowledge of these diverging views is critical to guiding the global shift to clean, affordable, and socially-just energy. Are individuals aware of the risks and benefits of the move to clean power? Do they believe their personal choices and behaviors will have an impact? Who do they feel should pay the costs of the transition? The World Energy Council commissioned Omdena to explore the effectiveness of artificial intelligence to grasp the attitudes of the world’s populace on these topics.

For this eight-week machine learning project, the team built numerous AI models to perform natural language processing. Known as NLP, this approach to AI is concerned with understanding human language. Social media conversations and news articles addressing energy-related topics served as the data for the project. The NLP models were trained to gather and categorize public conversations about energy transition topics. In the words of Amardeep Singh:

What sets this challenge apart from the rest was the sheer scale of data collected, social channels scraped and data analyzed.

For example, one set of models gathered and analyzed tweets in more than 20 countries that were related to complaints about “renewable energy cost”.

Figure 1: Renewable energy cost complaints

As seen in the chart, the modeling revealed that technology is the biggest concern in the complaint tweets in Brazil and France. In contrast, relevant tweets in Nigeria were focused solely on policy. Though conclusions cannot be drawn from these isolated collections of data, this exploratory work has led to an understanding of the boundaries of what can be extracted from public online sources. Omdena Collaborator Mahzad Khoshlessan applied various models to filter for relevant tweets to visualize thoughts, concerns, and sentiments of citizens in the USA, UK, Nigeria, and India. Below is an example visualization displaying the most discussed topics in India.

Figure 2: Word Clouds (India)

“Here at the World Energy Council, we recognize the opportunity and urgent need to humanize energy transition. Only by working at the human-level, embracing a broader community and addressing the social impacts agenda, will it be possible to achieve and sustain the breakthrough performance required for fast, clean, just, and socially inclusive global energy transition.” — The World Energy Council