LOTI Data Science Network: Extracting vulnerability data from free text

In Hackney's housing team, a lot of vulnerability information, such as safeguarding and support needs, is buried across more than a million free-text case notes, which means staff often have to review notes manually. 

This can lead to data gaps and means residents may have to repeatedly disclose difficult or traumatic circumstances to different teams. It also makes it difficult to proactively support residents or use this information for data-driven decisions.

A colleague at Hackney has developed a custom machine learning pipeline using named Entity Recognition, Relation Extraction and Entity Linking to identify additional needs in case notes and link them to the relevant residents. The aim is to help Housing Officers identify and review potential vulnerabilities, while also making this information more usable for migration and analysis.
About LOTI

LOTI is London local government’s collaborative innovation team. We help London borough councils and the GLA use innovation, data and technology to be high performing organisations, improve services and tackle London’s biggest challenges together. For more information, visit our website

If you have any issues registering for a ticket kindly email contact@loti.london