US and Singapore-based The Normal Company worked on two projects to classify online content to provide sustainability advisory for small businesses.

The problem

Local small businesses (e.g. gastronomy, retail, service providers) form an important part of our global economy as they employ 60% of the human workforce and generate 50% of the global GDP. But they are also responsible for 50% of all pollution and emissions. To achieve the 2030 UN SDGs, it is imperative that these 350 million+ businesses modernize and decarbonize their operation now.

To help these small businesses, Normal explored the following topics with Omdena: 

• How can we continuously source, classify and process the existing knowledge of what small business owners could and should be doing?
• How can we work with the resulting data to drive sustainable behavior change within the small business community?

The project outcomes

The team went from data collection to AI modeling to build, train, and fine-tune a machine learning model that classifies the content based on the following criteria: 

• Author/Authority: Reference, source, and domain authority
• Subject: Keywords and language
• Relevancy: Freshness of publication, and Impact/Sustainability relevancy.
• Format: Mood, Technicality, and depth

The results from this project will be further enhanced by putting a human in the loop to ensure that content is being accurately classified.

Partner testimonial