Natural Language Processing (NLP) is a subfield of AI that aims to help machines make sense of how humans write and speak. This is made particularly difficult by the fact that language by nature can be messy and inconsistent. In an effort to standardize text a subfield of NLP called Named Entity Recognition (NER) sprang up to extract and classify elements of unstructured text into a structured dataset to better help categorize and handle free text.
For example, the sentence “Jeremy is going to New York City on Thursday to attend Stern's Gala on American Business.” contains at least six different entities within it. We have the subject’s name (Jeremy), a date (Thursday), and a geopolitical entity (American) to name a few. Entity extraction, the act of identifying and classifying these elements of text, has long been a vital part of NLP.
Customer support organizations use entity extraction to route customer queries to the proper support center. Financial firms extract entities from news articles and financial documents to quickly sort incoming news and information. The applications of entity extraction are virtually limitless but the technology has up until now been for the very technical minded to set up and use.
That’s why we created Forefront Extract, an easy to use entity extraction API. Extracting high-quality entities from text of any length is now only an API call away. Using the API is easy and only requires two inputs:
"locations" - references to geographical locations like New York, or Istanbul
"people" - references to people generally by their name
"events" - references to a planned occasion like a meeting or a Gala
"organizations" - references to groups of people like Apple or a local food bank
"dates" - references to times like Thursday or simply “tomorrow”
"geopolitics" - references to large political entities like Americans or a government
Let’s see the Extract API in action.
Forefront Extract can take a free text query from a customer and provide the necessary tags that a routing system can use to assign the proper agent to the customer.
We can see that with Forefront Extract, a routing system can automatically recognize a customer’s location and send them to the proper support center with structured information so now the new agent will know the customer’s name and where they live.
The constant influx of financial news can be overwhelming but with Forefront Extract, we can extract the relevant entities from news articles to sort which news is the most urgent. Let’s use Forefront Extract to extract organizations from one news headline that is about Target and one that could have been confusing with just a keyword search.
We can see that the Extract API is smart enough to recognize that the first headline is about the organization “Target” whereas the second headline merely uses the capitalized word in context with the F.B.I.
It’s clear from our examples that Forefront Extract can intelligently extract entities from all kinds of text from multiple domains. The possibilities with what we can do with Forefront Extract are boundless. Documentation for the Extract API can be found here.
Ready to sign up and get started? Sign up for Forefront today!
Start fine-tuning and deploying language models or explore Forefront Solutions.