id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_nu5xpentbrcxdlcdt3lmatbur4	Vivek Srikumar	Modeling Semantic Relations Expressed by Prepositions	2013	12	.pdf	application/pdf	7668	606	56	This paper introduces the problem of predicting semantic relations expressed by prepositions and develops statistical learning models model the preposition relation and its arguments along with their semantic types, as a This paper addresses the problem of predicting semantic relations conveyed by prepositions in text. to model the predicate (i.e. the preposition relation) and its arguments (i.e. the governor and object). types to predict the preposition relation label. their types jointly and propose a learning algorithm that learns to predict all three using training data that annotates only relation labels. To aid better generalization and to reduce the label complexity, we follow this line of work to define a set of relation labels which abstract word senses across prepositions2. that our training data only annotates the relation labels and not the arguments and types. The first model aims at predicting only the relation labels and not the arguments and types.	./cache/work_nu5xpentbrcxdlcdt3lmatbur4.pdf	./txt/work_nu5xpentbrcxdlcdt3lmatbur4.txt