tailieunhanh - A memory–based learning approach to event extraction in biomedical texts

We do some initial processing of the raw-format mes- sages before the next step. The rst is to extract a reli- able sender IP address heuristically for each message. Al- though the message format dictates a chain of relaying IP addresses in each message, a malicious relay can easily al- ter that. Therefore we cannot simply take the rst IP in the chain. Instead, our method is as follows (similar to the one in [5]). First we trust the sender IP reported by Hot- mail in the Received headers, and if the previous relay IP address (before any server from Hotmail) is on our trust list (. other well-known mail. | A memory-based learning approach to event extraction in biomedical texts Roser Morante Vincent Van Asch Walter Daelemans CNTS - Language Technology Group University of Antwerp Prinsstraat 13 B-2000 Antwerpen Belgium @ Abstract In this paper we describe the memory-based machine learning system that we submitted to the BioNLP Shared Task on Event Extraction. We modeled the event extraction task using an approach that has been previously applied to other natural language processing tasks like semantic role labeling or negation scope finding. The results obtained by our system F-score in Task 1 and in Task 2 suggest that the approach and the system need further adaptation to the complexity involved in extracting biomedical events. 1 Introduction In this paper we describe the memory-based machine learning system that we submitted to the BioNLP shared task on event extraction1. The system operates in three phases. In the first phase event triggers and entities other than proteins are detected. In the second phase event participants and arguments are identified. In the third phase postprocessing heuristics select the best frame for each event. Memory-based language processing Daelemans and van den Bosch 2005 is based on the idea that NLP problems can be solved by reuse of solved examples of the problem stored in memory. Given a new problem the most similar examples are retrieved and a solution is extrapolated from them. As language processing tasks typically involve many 1Web page http . GENIA SharedTask subregularities and pockets of exceptions it has been argued that memory-based learning is at an advantage in solving these highly disjunctive learning problems compared to more eager learning that abstract from the examples as the latter eliminates not only noise but also potentially useful exceptions Daelemans et al. 1999 . The BioNLP Shared Task 2009 takes a .

• Tổ chức sự kiện
• Calculating Capital
• Delayed Exchanges
• Spanish Population
• Presentation
• Incidence and Survival
• Functionality
• Capital markets
• Lecture Money and capital markets
• Financial institutions
• Global marketplace
• Financial system
• Calculating interest rates
• Financial management
• Fundamentals of financial management
• Financial statement
• Quản lý tài chính
• Capital budgeting
• Calculating payback
• Finance management
• Fundamentals of finance management
• Discounted payback period
• Financial calculator solution
• Money markets
• Financial instruments