Texting has revolutionized the way humans interact on a personal and a professional level with one another. Now, the door to interacting with media organizations such as the Dallas Morning News through messaging has been opened through artificial intelligence.
UT’s journalism and computer science students have joined forces this semester to produce a chatbot for the Dallas Morning News entertainment site GuideLive that will answer questions, including what to do and where to eat, via Facebook Messenger based on GuideLive’s database and user preferences. Robert Quigley, journalism senior lecturer and Innovation Director for the School of Journalism, said the idea of a chatbot was first suggested to him by Vox Media COO Trei Brundrett and will come to life as a product of the “test kitchen” element of the innovation endowment from the Dallas Morning News for the School of Journalism.
“Bots have become a hot thing for people who are experimenting with new things in journalism,” Quigley said. “Voice bots have been around for a while now, but text bots where you can actually just talk to a bot through text is a relatively new idea — it has only been around for a little over a year.”
Dallas Morning News managing editor Robyn Tomlin said this new form of integrating technology with journalism seemed like the perfect way to attract a new audience and innovate what they already had.
“We love working with the folks at the innovation program here,” Tomlin said. “There is so much cool stuff happening with chatbots right now, and we wanted to see what that would look like for us. We thought, ‘Here is an opportunity to take (this website) and give it another life and see how people engage with that.’”
Graduate journalism student Kelsey Whipple said the chatbot will learn readers’ preferences based on the user’s Facebook profile and the questions posed by the user. She said based on demographic data of the Dallas Morning News’ GuideLive users, they decided to give the chatbot, named “Lucy” after the “Dallas” TV show character, the personality of their most common audience — a trendy 25-year-old woman.
“We want to make it easier for people to interact with the Morning News,” Whipple said. “So, people who are more mobile-friendly instead of desktop, or they only want events or they are on the go and don’t have time to find a computer, can just have a quick conversation with their thumbs and learn more about Dallas.”
Journalism senior Kylie Badgley, who is designing the chatbot’s personality, said she surveyed the typical responses of women in this age group and researched youthful quotes on Pinterest.
“I don’t really know how to sound trendy,” Badgley said. “But I made it sound like a 25-year-old. One example we were talking about today is if you type in ‘I want to go dancing,’ it will say something like, ‘Why be moody when you can shake your booty.’”
Badgley said this experience helped her learn a new aspect of journalism and gain a wider perspective on the work others do at UT.
“It has broadened my horizons,” Badgley said.
Although they have yet to discover the full potential for the chatbot, Tomlin said she already feels proud of all that the students have accomplished in its creation.
“We were really excited with what we saw,” Tomlin said. “It’s still not 100 percent done, but I think they have built something that is very innovative, smart and funny. We are really excited to take that and figure out what other things it needs to be successful in our market.”
