AI sports commentator knows all the best stories

General Discussions
Post Reply
User avatar
Nipuna
Moderator
Moderator
Posts: 2729
Joined: Mon Jan 04, 2010 8:02 pm
Location: Deraniyagala,SRI LANKA

AI sports commentator knows all the best stories

Post by Nipuna » Mon Oct 08, 2012 7:49 am

They think it's all over (Image: Getty Images/Flickr/rolfo)
They think it's all over (Image: Getty Images/Flickr/rolfo)
mg21628856.300-1_300[1].jpg (10.1 KiB) Viewed 2040 times
A database of relevant anecdotes for commentators during live sport means that the best stories are always to hand

WHEN watching sport on TV, the footage is often only half the fun. A good commentator can make all the difference, peppering a play-by-play account of the action with expert knowledge and anecdotes. But even the best commentator's repertoire is limited. Now an AI system called Scores could help.

Created by sports fan Greg Lee at the University of Alberta in Canada, Scores can tap into a stash of sporting stories to find relevant anecdotes that a commentator might not have thought of.

The idea came to Lee when he found himself choosing between matches on TV based on the commentary rather than the teams. Good commentators insert stories, he says. "As a viewer I was wanting them to tell me something interesting." The approach could also provide believable commentary to sports video games.

He and his team initially experimented with baseball commentary. During a match in which one team is trailing by four runs in the ninth inning, for example, the system might suggest an anecdote about the Los Angeles Dodgers, who came back to tie a game from the same situation in 2006 by hitting four consecutive home runs.

The system works by matching the features of a live event - such as the teams, key players, the score and the remaining time - against a database of available stories. Once stories that include some of those features are found it selects the few that are most relevant and suggests them to a human commentator.

The challenge is in evaluating the relevance of candidate stories and ranking them. Lee's system uses machine-learning techniques to do this. The most important feature was the teams involved and the second was the difference in number of runs.

To test the system, the researchers used it to create commentary for pre-recorded sports broadcasts and presented them to an audience of 16 volunteers, who said they found the commentary relevant and enjoyable. The team will present the work at the AIIDE conference in California this month.

Tom Rennie, a commentator and editor at UK radio station talkSPORT, has reservations about an overdependence on historical anecdote. "Too much commentary gets bogged down in stats," he says. "I prefer it when the focus is on what's going on." But he concedes such a system could help with sports that last all day, like cricket or baseball. "When there's 8 to 10 hours of commentary, stories can make it more interesting."
Post Reply

Return to “General Discussions”