Entities are additional elements of structured data which can be extracted from within the verbatims in your dataset. Entities include data points such as monetary quantities, dates, currency codes, organisations, people, email addresses, URLs, as well as many other industry specific categories (see below for an example).
Example email verbatim with monetary quantity and value date entities predicted and a trade ID entity
Unlike labels, the platform is able to predict entities as soon as they are enabled, as it can identify them based on their typical, or in some instances very specific, format and a training set of similar entities. Like labels, users are able to accept or reject entities that are correctly or incorrectly predicted, enhancing the model’s ability to identify them in future.
There are two kinds of entities within Re:infer. Some entities (such as ‘Organisation’ and ‘Person’) can be trained live in the platform by users, much like you would train a label. This means that accepting, rejecting or applying these entities will result in real-time updates for their predictions in that dataset. We are currently in the process of significantly expanding our capability of supporting live trainable entities.
Most of the entities available in the platform are pre-trained by Re:infer, and are further updated and refined offline using in-platform feedback provided by users. It’s still important for users to accept or reject these entities when reviewing verbatims, as the training signals they provide will be used to improve the platform’s understanding of that entity in future.