The company is a stalwarts in the research and analysis of climate strategies. They synthesize the data of climate information and derive insights for the companies driving climate change within their org.
The company wanted to derive insights from the environmental, social and governance data references in the financial filing statements of the Fortune 100 companies. The data downloaded from the filing sites required extraction and cleansing before being ingested by machine learning software for text mining.
A master of key tags was created for the ESG data across various companies across industries, based on domain knowledge about ESG.
Various text mining algorithms were applied on the cleansed data for creation of tags relevant to ESG. Algorithms were also used to map these data tags with the key master and analyze the frequencies of the same.
Patterns evolving from these tag occurrences and clusters arrived at were noted and used for further analysis.
Bridgera had been the preferred software vendor for this customer for some time. Our team had a good understanding of the key needs of the customer as well as the expertise in the areas of business intelligence, analytics, text mining, and developing BI solutions to provide a proof of concept for the text mining solution to the customer.
- Word clouds for frequency of occurrences of the ESG tags present in the data
- Distinct clusters of the group of ESG tags present in the data, suggestive of joint occurrence of these tags across various companies
- Derivation of topics of importance across companies
Bag of Words
Bag of words for the frequency of tags
Hierarchical cluster analysis for grouping of tags
Neural networks for derivation of topics of importance
Multidimensional scaling for seeing the similarities and differences between tags
Data visualization using clustering and word clouds