The Australian Securities and Investments Commission (ASIC) wants to explore whether natural language processing is suited to help overcome the regulatory problems the government agency has, as well as if the tech would benefit or aid decision-making.
ASIC is seeking expressions of interest from vendors keen to conduct pilots of their natural language processing technologies and approaches, with the trials expected to put the corporate regulator in a position to determine if it should further invest in the area.
It is expected the pilot would also see staff understanding how the technology works, with an ancillary objective to assist ASIC in developing in-house skills and expertise in the area.
In preparing its approach to market (ATM), ASIC highlighted six "problem" areas it thinks natural language processing may help to solve, with the first helping it sort through social media and other advertising mediums -- both offline and online -- to find instances of misleading representations in the finance sector.
It also wants to speed up the time it takes to sift through the sale of insurance, which it currently does by listening in to phone calls.
"Depending on the available technology, this could mean converting voice-to-text sales calls then analysing the text transcripts or preferably a one-step process that does both at once," ASIC wrote.
"ASIC would expect that the technology would need to take into account nuances of speech such as tone and timing, and patterns of dialogue, to pick up potential issues such as consumer hesitation, excessively pushy sales techniques, and/or consumer confusion and misunderstanding."
Similarly, ASIC wants to determine if automated data reviewing capabilities would help it work out if managed fund product disclosure statements and financial advice was delivered to customers in a fair manner. It is also keen to have a bot search through financial reports and audits, as well as fundraising documentation, rather than having its staff manually read every item that is filed.
ASIC recently turned to the market for the provision and management of a data analytics and reporting tool, seeking a central location for its Strategic Intelligence team to find and analyse financial and economic data on both domestic and international organisations.
The watchdog is in the process of rolling out its 2017-20 data strategy, which outlines how ASIC will capture, share, and use data as it attempts to transition to a more "data-driven and intelligence-led organisation".
"Our data strategy is part of 'One ASIC' which is about 'connecting the dots' to achieve better regulatory outcomes," the strategy explains. "It is about working together and sharing data seamlessly using common language, systems, and processes."
To make this happen, ASIC will be establishing a repository that brings together the captured and created regulatory data. It will also be implementing an integrated CRM system to replace many legacy workflow systems, as well as developing a new online portal.
At the time, ASIC said it is also keen to explore options for receiving and accessing data, including through the use of regulatory nodes using distributed ledger technology, blockchain.
The natural language processing ATM closes March 19, 2018.
PREVIOUS AND RELATED COVERAGE
The government-owned corporate, markets, and financial services regulator has extended its digital investigation and analytics software contract with Sydney-based Nuix.
The Australian Securities and Investments Commission's new data strategy will see the regulator transition to a more 'data-driven' organisation that is keen to explore technologies such as blockchain, it has said.
After building a business case and working alongside the ATO to develop the platform, the department decided not to go live with its virtual assistant pilot.
Will 2018 be the year of the chatbot? Not without human help (TechRepublic)
Chatbots are still in the early stages, analysts say, but knowing how and when to use them in the enterprise may be a catalyst for their growth.
Top 5: Things AI might actually be good for (TechRepublic)
It's easy to dismiss AI as a buzz phrase or just a marketing term, and in some cases it is. But here are five situations where AI can be helpful.