Developing a better computer translator

NPR : To stay on top of what’s going on in the Middle East and South Asia, US political scientists, historians, and even the Pentagon are studying social media such as Facebook, Twitter, and blog posts. Because the posts are written in the local languages, computer translation programs that can not only translate the words but also convey the subtler nuances of sentiment and opinion are being researched. Funded by the Pentagon, a computer scientist at the University of Buffalo in New York, Rohini Srihari, has developed a natural language program that has “learned” the nuances of Urdu—a mix of Hindi and Persian that uses Arabic script and is spoken in Pakistan. According to Srihari, the program can extract factual information as well as positive or negative sentiment and opinion. Although Srihari acknowledges that the program is not perfect, she has gained “insight into what Urdu speakers have been talking about lately,” according to Christopher Joyce reporting for NPR.