Can artificial intelligence improve service and efficiency, and maybe even cut costs, in the public sector?

August 14, 2009 at 1:54 pm 1 comment

By Peter Baeck

Yesterday I suggested that government agencies such as HM Revenue & Customs should create an online community, where the users of the agency’s services could discuss the service level, as well as suggest new and maybe more efficient ways for the agency to improve its practice. Why hasn’t this happened? One answer could be that most public sector agencies are not to keen on their possible flaws being exposed to the masses, and this least of all in a online community the agency gives legitimacy to by hosting it. (More on why I disagree with this approach in next week’s blog.)

Another explanation to why public agencies aren’t setting up online communities is that the question still remains how you deduct knowledge from these possibly very large communities (NHS community for example, could potentially have millions of users posting millions of threads everyday), and convert it in to something that can benefit the organization, generate new ideas and thereby hopefully improve service efficiency and quality. Simple ways could be following online user polling, following the most popular and most commented threat, but these methods wouldn’t cover the whole of the community and could potentially take up significant resources. An alternative solution could be to introduce the artificial intelligence (AI) in to the monitoring of these communities. AI computer programs could potentially learn to scan a community, such as a hypothetical NHS, and tell policy makers, managers at all levels etc, which are the most often mentioned departments, diseases, are they positive or negative etc. Information that could potentially improve efficiency significantly.  There is some experience with this practice in other countries; one is the Cornell e-Rulemaking Initiative described below.


“To help the agencies deal with rule-making in the Internet age and make the process more accessible to the public, Cornell scientists and legal experts have created the Cornell e-Rulemaking Initiative (CeRI), funded by a $750,000, three-year grant from the National Science Foundation…. The potential of the Internet has some regulators worried. A proposed rule about organic-food labeling, for example, generated some 400,000 e-mail comments. A Federal Communications Commission rule about the consolidation of media ownership brought in over 2 million comments. Such extremes have been rare, but since regulators are required to consider every submission and respond to major points, they are asking for help.Claire Cardie, an expert in natural language processing, is developing computer programs to sift and categorize the masses of comments. First, agency staff will highlight sentences in the comments that connect with various issues. Over time, the computer will learn the rules of classification and take over.’People can classify all of the phrases and sentences in about 40 to 50 comments per day, depending on length,’ Cardie says. ‘Software takes just seconds to classify all of the phrases and sentences in one document.'” Cornell Chronicle Online (May 16, 2007).

So what do you think, can artificial intelligence improve service and efficiency and maybe even cut costs in the public sector? 


Entry filed under: Government Departments, Innovation Policy, Research.

An efficient way to get a National Insurance Number please!! Innovation here and there!!

1 Comment Add your own

  • 1. Claire McEneaney  |  August 14, 2009 at 2:39 pm

    I like the potential of this idea. Its been a while since I studied AI so I’m not sure how the algorithms have advanced! They can be extremely complex and from what I know of before they can often misinterpret data (particularly when people use sarcasm!).

    It would require some very good mechanisms to ensure it works and I don’t think you can escape the fact that it would need some kind of human input e.g. to interpret the outputs.

    Also, would public sector agencies find this information useful and innovative, or would it just confirm what they already know about their strengths and weaknesses?


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

Trackback this post  |  Subscribe to the comments via RSS Feed

The Innovation Unit website

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 29 other followers


Twitter Updates

Follow innovation_unit on Twitter

%d bloggers like this: