Enriching Funding Application data using Named Entity Recognition


Entity extraction


The Wellcome Trust wanted to get more insight about where the benefits are being felt of the grants that they fund, to better inform decisions about which grants to fund.


The Wellcome Trust (Wellcome) stores over 130,000 applications for funding. Much information is captured in free text fields, such as the title, and summary. These free text fields are not easy to analyze, but we can use machine learning to extract insights that would otherwise be very manual and time consuming.


We extracted locations from free-text fields in grant applications, allowing Wellcome to analyse these, and integrated the solution into their existing infrastructure so that the analysis can be repeated frequently and automatically.


The work led to a 35% increase in the amount of location information that could be extracted from grant applications, and is helping Wellcome to answer: who are the ultimate beneficiaries of research funding.

You can read more about this case study on our blog or in more details at the Wellcome Trust Blog. We’ve also created an example application that you can try for yourself.

Do you have a Natural Language Processing problem you need help with?

Let's Talk