| 引用本文: | 田野,王文东,饶京海,王冠,郭亮,陈灿峰,马建.短信息的会话检测及组织.软件学报,2012,23(10):2586-2599 |
| |
|
| 本文已被：浏览 4858次 下载 7727次 |
码上扫一扫！ |
|
|
| 短信息的会话检测及组织 |
|
田野1, 王文东1, 饶京海2, 王冠1, 郭亮1, 陈灿峰2, 马建1,3
|
|
1.网络与交换国家重点实验室(北京邮电大学), 北京 100876;2.诺基亚研究院, 北京 100176;3.无锡物联网产业研究院, 江苏无锡 214135
|
|
| 摘要: |
| 如何挖掘存储在手机上的大量短信息背后所隐含的会话信息,是一个非常具有挑战性的问题,因为它们并不具备“主题”、“回复”等经常被用于邮件线索分析的元数据.基于此,提出了一种基于时间聚类算法和话题检测的短信息会话识别模型.首先,根据短信息流的时间分布特性,将会话双方的所有短信息划分到一个一个的候选会话中,进而运用基于 latent Dirichlet allocation(LDA)训练出来的语义话题模型,对候选会话进行更深层次的分析;利用该话题模型度量了各个候选会话在话题上的相关度.最后,在综合时间和话题相关度的基础上,通过对候选会话的合并识别出隐含的会话信息.通过对包含了 50 名大学生在 6 个月中产生的 122 359 条短信进行实验验证,证明了该算法的有效性. |
| 关键词: 短信息 时间聚类 话题 latent Dirichletallocation |
| DOI：10.3724/SP.J.1001.2012.04191 |
| 分类号: |
| 基金项目:国家重点基础研究发展计划(973)(2009CB320504); 国家高技术研究发展计划(863)(2011AA01A101) |
|
| Conversation Detection and Organization of Mobile Text Messages |
|
TIAN Ye1, WANG Wen-Dong1, RAO Jing-Hai2, WANG Guan1, GUO Liang1, CHEN Can-Feng2, MAJian1,3
|
|
1.State Key Laboratory of Networking and Switching, Beijing University of Posts and Telecommunications, Beijing 100876, China;2.Nokia Research Center, Beijing 100176, China;3.Wuxi SensingNet Industrialization Research Institute, Wuxi 214135, China
|
| Abstract: |
| Mining the latent conversations which are implied in the big amount of text messages stored on one’smobile phone, is a challenging problem. They can hardly be organized by threads, due to lack of necessary metadatasuch as “subject” and “reply-to”. This paper proposes an innovative conversation recognition model based ontemporal clustering algorithms and topic detection methods. The study first clusters the text messages into candidateconversations based on their temporal attributes, and then does further analysis using a semantic model based onlatent Dirichlet allocation (LDA). In the end, the text messages are organized as conversations based on theirintegrated correlation of temporal relevancy and topic relevancy. This approach is evaluated with a real dataset,which contain 122 359 text messages collected from 50 university students during 6 months. |
| Key words: text message temporal clustering topic latent Dirichlet allocation |
