Topic Detection about People from Blogsphere by Using Mnemonic Names
Tomoko Hokama, Hiroyuki Kitagawa · 2007
People are referred to by various ways in weblog articles. For example, a person may be referred to by the full name, affiliation and title, or nicknames. These mnemonic names reflect writers’ images or opinions for the person. In this paper, we extract mnemonic names for the target person. Then we collect weblog articles that include the full name and weblog articles that include each mnemonic name separately. After collecting weblog articles, we extract topics from each article set and analyse differences between extracted topics from each article set.