By Serla Rusli City University of London London, United Kingdom
Matthew Gould, chief strategy officer of Arria, spoke to the audience at the 2017 Big Data Media Week conference in London on Artificial Intelligence (AI) and the power of language, explaining how the UK-based company offers AI technology in data analytics and information delivery using Natural Language Generation (NLG) technology.
Introducing the subject through the concept of anthropology, Gould said, “Humans are the only species with the ability to have unstructured environmental inputs, record it, and convey it.”
The challenge in the AI area is thus to replicate this ability — having data coming in and being understood in context. Below are some of the key aspects of AI to be able to mimic human capabilities and scale them:
Data ingestion: Data can be ingested from a wide variety of sources both structured and unstructured. This can be further broken down into:
- Data analysis: processes the data to extract the key facts that it contains.
- Data interpretation: Makes sense of the data, particularly form the point of view of what information can be communicated.
Presentation: Narrative can be output in a variety of formats and combined with graphics or delivered as speech. This includes:
- Document orchestration: Takes the messages derived from data and structures the information they contain into a narrative.
- Microplanning: Works out how to package the information into sentences to maximise fluency and coherence.
- Surface realisation: Ensures that the meanings expressed in the sentences are conveyed using correct grammar, word choice, morphology, and punctuation.
In the media context, Arria aims to automate the ability to gain expertise.
“Expertise can only be communicated through language,” Gould said. “What we want to do is to provide content producers with the ability to combine data with the expertise, and allow the producer to have a choice on how the data are being communicated.”
The first benefit of using AI in media is in increasing the volume of content produced, Gould said. Often, even the simplest media content — such as the weather report — still needs human explanation. AI will be able to automate many of these processes and increase the volume of content being produced. The UK Met Office could previously only produce 60 human-authored weather reports in 24 hours; with the help of Arria’s Platform, the office can produce 5,000 reports in the same amount of time.
The second benefit is on the ability to tailor messages: “It’s almost rude when we don’t tailor communication with other people,” Gould said.
With Big Data and AI, it is possible to put expertise into the system and communicate data to audiences in an engaging way.
Serla Rusli is a business and financial journalism MA student at City, University of London. She can be reached email@example.com.