Exploring natural language processing with Croesus

While looking to the future of FinTech, those in the Croesus Lab have found themselves enamored with natural language processing, and the implications it has for their field

Exploring natural language processing with Croesus

The Croesus Lab has a number of projects on the go, but one in particular has real world applications: natural language processing (NLP).

NLP is concerned with the way computers interact with human language, particularly how to program computers to process as well as analyze large amounts of natural language data, and the researchers in the Croesus Lab are looking to expand on NLP in the FinTech space.

The lab itself looks to innovate in the world of FinTech by researching projects, such as NLP, artificial intelligence, and security. It is made up of three sections: research teams, a group which promotes innovative culture, and a sector which helps startups to get their feet off the ground.

According to Maxime Dumas, the principal researcher of technological innovation at Croesus, they are doing a lot of work in NLP, with a lot of effort being put into summarization.

“We want to be able to find out more about summarization, question answering, and information retrieval. All of these are really interesting for us. We’re trying to simplify the work, or reduce the time you spend on several things,” he explained.

One example he gives is summarizing the news. “That would be awesome,” he said. “We’d be able to give you the answer. So, whenever you have a long text to read, we’d be able to summarize it and give you some insights. That way you’ll spend less time reading and going through everything.”

Ultimately, they’re trying to target information by finding the answer to what one is looking for and finding the short answer for it.

According to Patrick Chasse, the director of research and development at Croesus, technology which uses NLP doesn’t always understand what people are looking for, so, “sometimes it’s painful.” But, he noted there is hope, as this technology advances very quickly.

“So, in a couple of years, maybe it will be a completely different picture there,” he said.

At this point, the NLP team is focusing on extracting insights from unstructured documents. There is a gold mine of information hidden in text files.

“Structured or unstructured data that you could query and get some information out of it. That’s more of what we’re targeting,” said Dumas.

Larger, generic tasks around natural language, such as translation, text-to-speech or speech-to-text, are problems being solved by companies such as Google, Microsoft, and MIT. At Croesus, Dumas explained they are looking at industry specific problems.

“We have one in-house researcher per project, and a few students. So, we can’t do work like Google which spends millions on those problems,” he said, adding what they’re trying to tack precisely is niche problems in technology, and finding out how they can add a little bit to a larger piece of technology.

In the end, the Croesus Lab is looking to “push boundaries,” said Chasse and Dumas, as each project they do looks to innovate, grow, and propel the company toward the future of the FinTech industry.

To find out more about Croesus, and the Croesus Lab, visit their website at croesus.com