stub Researchers Create System to Extract Info From COVID-19 Articles - Unite.AI
Connect with us

Artificial Intelligence

Researchers Create System to Extract Info From COVID-19 Articles

Updated on

A group of researchers at the University of the Basque Country has developed a prototype for VIGICOVID, which is an automatic information extraction system for COVID-19 scientific articles. The system relies on natural language questions to get answers regarding COVID-19. 

The project involving VIGICOVID was run by the UPV/EHU’s HiTZ Centre, the UNED’s NLP & IR Group, and Elhuyar’s Artificial Intelligence and Languages Technologies Unit. 

The research was published in journal Knowledge-Based Systems.

Changing the Information Search Paradigm

Eneko Agirre is head of the UPV/EHU’s HiTZ Centre.

“The information search paradigm is changing thanks to artificial intelligence,” said Agirre. “Until now, when searching for information on the internet, a question is entered, and the answer has to be sought in the documents displayed by the system. However, in line with the new paradigm, systems that provide the answer directly without any need to read the whole document are becoming more and more widespread.”

Xabier Saralegi is an Elhuyar researcher.

“The user does not request information using keywords, but asks a question directly.” 

The system relies on two separate steps to search for answers. 

“Firstly, it retrieves documents that may contain the answer to the question asked by using a technology that combines keywords with direct questions. That is why we have explored neural architectures,” Dr. Saralegi said. 

Deep Neural Architecture

The team relied on deep neural architecture fed with examples. 

“That means that search models and question answering models are trained by means of deep machine learning,” he continued. 

The set of documents is first extracted before being reprocessed through a question and answer system, which helps obtain specific answers. 

“We have built the engine that answers the questions; when the engine is given a question and a document, it is able to detect whether or not the answer is in the document, and if it is, it tells us exactly where it is,” Dr. Agirre said. 

According to the researchers, they were happy with the results of their work.  

“From the techniques and evaluations we analysed in our experiments, we took those that give the prototype the best results,” Dr. Agirre continued. “We have come up with another way of running searches for whenever information is urgently needed, and this facilitates the information use process. On the research level, we have shown that the proposed technology works, and that the system provides good results.” 

“Our result is a prototype of a basic research project. It is not a commercial product,” Saralegi added. 

With that said, these types of prototypes can be modeled in a short period of time, meaning it might not be long before there is a commercial product. 

Alex McFarland is a tech writer who covers the latest developments in artificial intelligence. He has worked with AI startups and publications across the globe.