Automating text mining analytics at a major philanthropic foundation
- Working in all 50 states and more than 100 countries, a multitude of partnerships, and working in a variety of unrelated, but worthy causes, it is clear that honing the grant selection process for success is imperative.
- Factors that contribute to grant success are difficult to determine and may seem unrelated.
- Identify factors indicative of failure or success using grant proposal data and external data sources, to increase the efficacy of grant funding.
- Using Microsoft Cognitive Services and Bot Framework, Neal Analytics developed a chatbot capable of asking novel questions, displaying on-the-fly data metrics, and giving insight.
- Behind the scenes, many different machine learning algorithms and text mining techniques were required.
- By combining both internal and external data sources, a high predictive analysis is now possible from the grant proposals alone.
- Bot outputs graphical and text-based insight
- Bot automates user queries and answers a key question
- Bot surfaces