Báo cáo khoa học: "Detecting Semantic Relations between Named Entities in Text Using Contextual Features" pdf

... Proceedings of the ACL 2007 Demo and Poster Sessions, pages 157–160, Prague, June 2007. c 2007 Association for Computational Linguistics Detecting Semantic Relations between Named Entities in Text Using ... learn- ing method for detecting a semantic rela- tion between a given pair of named enti- ties, which may be located in different sen- tences. The method employs n...

Ngày tải lên: 17/03/2014, 04:20

... our ap- proach to capture meaning equivalence and informa- tion disparity in cross-lingual texts. (b) In both settings the combination of lexical, syn- tactic and semantic features (PT+SPT+DR) ... or an “integrated solution” that exploits bilin- gual phrase tables to capture lexical relations and contextual information (Mehdad et al., 2011). The promising results achieved with the i...

Ngày tải lên: 19/02/2014, 19:20

... remainder is used in training. 4 Models Our starting point for statistical NER is a feature- based linear model over sequences, trained using the structured perceptron (Collins, 2002). 8 In addition ... with self-training on unlabeled target-domain data; enforc- ing the same recall-oriented bias in the self- training stage yields marginal gains. 1 1 Introduction This paper considers...

Ngày tải lên: 24/03/2014, 03:20

... conceptual in- stances is obtained by deleting from each candi- date those instances that are subsumed by a higher scoring conceptual instance. Remaining conceptual instances are then re-scored using ... ontologizing binary semantic relations into WordNet and we present empirical results on the task of attaching part-of and causa- tion relations. Formally, given an instance (x, r...

Ngày tải lên: 20/02/2014, 12:20

... for training and testing. The combined total number of sentential contexts for our 453 NCs was 7,714, containing 1,165 distinct main verbs. We next randomly split the NC data into 80% training data ... number of training NCs varies depending on the verb-mapping method. As noted in Section 2, the relations TIME and EQUATIVE are not associated with seed verbs. For TIME, rather than using...

Ngày tải lên: 08/03/2014, 02:21

... binding re- lationships. In Srinivasan and Rindflesch (2002) MeSH term co-occurrences within MEDLINE ar- ticles are used to attempt to infer relationships be- tween different concepts, including ... late-breaking bioscience infor- mation is found only in textual form, and so meth- ods are needed to automatically extract semantic entities and the relations between them from this te...

Ngày tải lên: 17/03/2014, 06:20

... features. 1 Introduction Temporal information encoded in textual descrip- tions of events has been of interest since the early days of natural language processing. Lately, it has seen renewed interest ... simultane- ous. We map the thirteenth identity into simultane- ous. One oddity is that Timebank includes both dur- ing and included by relations, but during does not appear in Timebank...

Ngày tải lên: 23/03/2014, 18:20

... NIL queries in our training data. After normalizing fea- ture values to be in [0, 1], we train a candidate and a NIL classifier on 90% of the queries in each it- eration, and test using the remaining ... Christopher Manning. 2005. Incorporating non-local information into information extraction systems by gibbs sampling. In Proceedings of the 43rd Annual Meeting on Associ- ation for C...

Ngày tải lên: 23/03/2014, 16:20

... IFN-c, interferon-c; IL, interleukin; IP-10, interferon-c-inducible protein-10; MAPK, mitogen-activated protein kinase; MCP, monocyte chemoattractant protein; Mig, monokine induced by interferon-c; ... and IFN-c. (B) Chemokines expressed in the joint recruit leukocytes into the joints. In addition to functioning in cell trafficking, several chemokines have other biological abilities. Che...

Ngày tải lên: 18/02/2014, 18:20

... Training set size(words) Precision 25k 88.7% 200k 91.8% Number of Appearances in Training Set • • 4- - • Machine Students Figure 3: Test Set Recall vs. Frequency of Appearances in Training ... [DGNPZ] ?) The first line denotes the action, in this case, Add a bracketing. The second line defines the context preceding the sequence we want to have bracketed in this case, we do...

Ngày tải lên: 20/02/2014, 19:20