An Automatic Topic Summarization Using Content Anatomy

A topic is defined as a seminal event or activity along with all directly related events and activities. It is represented by a chronological sequence of documents published by different authors on the Internet. We define a task called topic anatomy, which summarizes and associates the core parts of a topic temporally so that readers can understand the content easily. The proposed topic anatomy model, called TSCAN, derives the major themes of a topic from the eigenvectors of a temporal block association matrix. Finally, the extracted events are associated through their temporal closeness and context similarity to form the evolution graph of the topic While current technologies are efficient in searching for appropriate documents to satisfy keyword search requests, users still have difficulty assimilating needed knowledge from the overwhelming number of documents. The situation is even more confusing if the desired knowledge is related to a temporal incident about which many independent authors have published documents based on various perspectives that, considered together, detail the development of the incident.