Đang chuẩn bị liên kết để tải về tài liệu:
A Markov Logic Approach to Bio-Molecular Event Extraction

We successfully found hundreds of botnets by examin- ing a subset of the spam email messages received by Hot- mail Web mail service. The sizes of the botnets we found range from tens of hosts to more than ten thousand hosts. Our measurement results will be useful in several ways. First, knowing the size and membership gives us a bet- ter understanding on the threat posed by botnets. Second, the membership and geographic locations are useful infor- mation for deployment of countermeasurement infrastruc- tures, such as rewall placement, traf c ltering policies, etc. Third, characterizing botnets behavior in monetiz- ing activities may help in ghting against botnets in these businesses, perhaps reduce their. | A Markov Logic Approach to Bio-Molecular Event Extraction Sebastian Riedel Hong-Woo ChunU Toshihisa Takagi Jun ichi TsujiiTS Database Center for Life Science Research Organization of Information and System Japan Department of Computer Science University of Tokyo Japan Department of Computational Biology University of Tokyo Japan School of Informatics University of Manchester UK National Centre for Text Mining UK sebastian chun takagi @dbcls.rois.ac.jp tsujii@is.s.u-tokyo.ac.jp Abstract In this paper we describe our entry to the BioNLP 2009 Shared Task regarding bio-molecular event extraction. Our work can be described by three design decisions 1 instead of building a pipeline using local classifier technology we design and learn a joint probabilistic model over events in a sentence 2 instead of developing specific inference and learning algorithms for our joint model we apply Markov Logic a general purpose Statistical Relation Learning language for this task 3 we represent events as relational structures over the tokens of a sentence as opposed to structures that explicitly mention abstract event entities. Our results are competitive we achieve the 4th best scores for task 1 in close range to the 3rd place and the best results for task 2 with a 13 percent point margin. 1 Introduction The continuing rapid development of the Internet makes it very easy to quickly access large amounts of data online. However it is impossible for a single human to read and comprehend a significant fraction of the available information. Genomics is not an exception with databases such as MEDLINE storing a vast amount of biomedical knowledge. A possible way to overcome this is information extraction IE based on natural language processing NLP techniques. One specific IE sub-task concerns the extraction of molecular events that are mentioned in biomedical literature. In order to drive forward research in this domain the BioNLP Shared task 2009 Kim et al. 2009 concerned the extraction of .