VIDEO | Moody’s Analytics Library Sciences Head: We Process Close To a Million News Stories a Day

(US and Canada) Jon Campbell, Head of Library Sciences at Moody’s Analytics, speaks with Mark Johnson, Regional VP and Editorial Board Chair of CDO Magazine, in a video interview about the NLP concept, the NewsEdge NLP Engine, its key benefits and various business use cases.

Campbell defines Natural Language Processing (NLP) as using computers to analyze text. He says that the goal for the NewsEdge team is to enrich news and textual data with additional signals such as industries, subjects, locations, essential people, sentiments, event classification, and news clustering.

Speaking on the need for the NewsEdge Natural Language Processing Engine, he explains that the solution enables decision making using news, and the core engine processes news quickly with a high-speed pipeline.

He shares that NewsEdge has had customers seeking solutions for specific use cases, such as news publishers wanting to categorize and tag content or algorithmic trading firms wanting to run the software in-house. Campbell adds that Moody’s Analytics, too, uses the NewsEdge NLP Engine to categorize its textual content and research reports.

He goes on to mention the following benefits of the NewsEdge NLP Engine:

• Active feedback loops with customers, QA staff, and editorial teams to constantly tune the models to improve them, analyze the outputs, and further improve the product.
• Complete transparency and explainability showcase why classifications or metadata tags were applied to the data. This drives fair business practices and governance.
• Fixed subscription with a flat fee allows users to use the tool as much as they require instead of paying per transaction.
• It’s a maintenance-free model. Customers don’t need to deploy additional personnel, such as software developers, data scientists, and editorial and library science staff.
• It can run entirely in-house without any external communication.

Elaborating further, Campbell says that the solution matches the high confidentiality standards established by certain businesses and government organizations. Companies can have their news feeds, emails, and research reports internally. Similarly, government security organizations can analyze large amounts of documents looking for security risks.

Next, Campbell showcases an internal tool that combines two independent information sources — a news story in the public domain and an internal email — to create relevant business information for a global construction company. Extracting entities from a news story and categorizing them pulls out information like competitor information, risk, locations, relevant topics, and geopolitical events.

He demonstrates how a public story is followed up with an internal email from someone who has insight into the topic. Campbell adds that processing the news and email in a combined manner leads to actionable results.

In conclusion, Campbell says that the engine processes nearly a million news stories daily and can categorize a story in under a second.

CDO Magazine appreciates Jon Campbell for sharing his insights and data success stories with our global community.