16 May 2014
Current Students/ Faculty & Staff/ Informal Learning Opportunities/ Journalism
Beyond Google: Is news personalization a new
job for librarians, journalists and technologists?
KINGSTON, R.I. – In a digital battle to personalize and curate news and the web for individual users, who wins – computer algorithms, people, or both?
As Internet access shifts to tablets and cells, will the fatigue of having to “order up” content via search will give way to smarter systems that will present content based on a user’s confirmed interest profile? These and other questions are on tap for a “brown-bag” entrepreneurial luncheon seminars, two experts will discuss developments in the field of artificial intelligence for editing and curation – and will show their experiment in the field during: “Beyond Google: New approaches to reach the Daily Me.”
DATE: Friday, May 16
TIME: 12 – 1:30 p.m.
LOCATION: Davis Hall, Seminar Room, URI Kingston Campus
Bring your own lunch; box lunches will be available for $5 each if you request one in advance by emailing firstname.lastname@example.org at least 24 hours in advance.
The two presenting entrepreneurs are:
- Graf Mouen, co-founder and CEO of Taxonometrics Inc., a New York-based development-stage company working on news personalization. Mouen formerly worked 15 years for ABC News as a technology project director and engineer.
- Bill Densmore, Taxonometrics co-founder and a research consultant to the Harrington School. Densmore is a former wire-service and newspaper reporter, editor, publisher, and Internet technology entrepreneur.
Densmore and Mouen will discuss the challenges of news personalization, and will demonstrate a service called “Newshare/Circulate” which combines computer algorithms with human topical curation. They will demonstrate the results – a constantly changing stream of news and information tailored to your expressed and inferred interests, and ask for feedback on how to improve it. Their work began as an experiment for the Donald W. Reynolds Journalism Institute at the University of Missouri.
“My years of work at ABC News has shown that computer algorithms alone can’t deliver perfect personalization,” says Mouen. “What we’re prototyping is a way of combining human curation of information with new algorithms. It requires the information science skills of librarians, the news-culling sense of journalists and the code-writing expertise of software engineers. We want to know if the results we’re getting are useful.”
“Knowledge discovery” is a large field and includes mainstream search engines (Google, Bing), alternative search engines, social media knowledge discovery tools (Facebook, Twitter), RSS readers, and news aggregators.
“We hope newspapers can adaopt Newshare/Circulate to offer a more complete and highly personalized service to their current and prospective users,” says Densmore. “On the mobile and fixed web, people are spending most of their time on entertainment applications, game applications, on social networking and searching – not on news or newspaper websites. As a result, the business model that supported journalism is falling apart.”
Such a process imitates the way that we discover people’s interests. Waiters and wine stewards, shop clerks, travel agents, matchmakers, even ordinary folks getting to know one another often make reasonable inferences “If you like A, then you must like B”. The ability to do this depends only locating A within a tree of possibilities and noticing what is nearby.