Finch for Text® Adds Text Summarization Capability

Reston, VA – Finch Computing, developers of the powerful natural language processing technology Finch for Text®, today announced that it has added text summarization as an available capability in the product.

“We know that, for many of our customers, the ability to quickly summarize documents or even entire corpuses of documents is a critical capability,” Finch Computing Chief Technology Officer Scott Lightner said. “This capability lets them do that at-scale on real-time or streaming content and gives them insights they could not glean previously, certainly not at this speed.”

Text Summarization involves creating short, fluent summaries of documents without losing key information or the document’s overall meaning and intent. Doing this accurately is challenging – doing it accurately, quickly and at scale proves even more so.

Presently, there are two main ways of approaching text summarization. The first is called extractive summarization, which involves selecting a subset of phrases or sentences in the original text and combining them to form a summary. The second type, abstractive summarization, involves generating phrases or sentences, possibly rephrasing them or using words that were not in the original text. Abstractive summarization provides more concise and human-like summaries but can be slow and inaccurate on longer documents. Additionally, redundancy becomes an issue when generating multi-sentence summaries.

“Finch Computing uses both abstractive and extractive summarization to produce better, more accurate summaries for our customers,” Lightner continued. “We leverage cutting-edge summarization models that combine both approaches and use reinforcement learning to ensure the two approaches work in harmony. We’ve trained those hybrid models on more than five years of news content. This intensive training on a large and broad corpus of text enables us to use the models on our customers’ text and to deliver accurate, high-quality results that our customers can trust.

Use cases for text summarization include summarizing news documents to identify trends; analyzing engineering reports to mitigate risk; assessing internal emails to identify insider threats or to gauge employee opinions; monitoring message boards and chat rooms; and identifying opportunities for cost savings and efficiencies – among others.

To learn more about Finch Computing, or about Finch for Text®, please visit finchai.com.

###