摘要
为了发现新闻故事中的关键说话人,用以提高多媒体检索效率,在说话人索引的基础上,提出了关键人发现方法:根据新闻故事中说话人的特点,基于说话人频率、说话人持续时间、平均每次说话人时长和说话人位置因子4个因素,综合定义了说话人关键度,用以判断说话人的重要性,把每个新闻故事中说话人关键度最大的人作为关键说话人。实验结果表明,该种算法可以找到故事中绝大部分的关键说话人,验证了该算法的有效性和可行性。
To solve the problem of key speaker discovery in News story and improve the efficiency of multimedia retrieval,on the basis of speaker index,the method of key speaker discovery is proposed.Speaker key is synthetically defined by speaker freque-ncy,speaker duration,average every time speaker length and speaker position factor,which is used to judge the speaker's importance and the biggest speaker key is regard as key speaker in every story.The experimental result shows that can find most key speaker in story and feasibility and effectiveness of the algorithm is demonstrated.
出处
《计算机工程与设计》
CSCD
北大核心
2012年第6期2353-2357,共5页
Computer Engineering and Design
基金
国家自然科学基金项目(61101160)
广东省自然科学博士启动基金项目(10451064101004651)
中央高校基本科研业务费专项基金项目(2011ZM0029)