Pontus Stenetorp

Learn More
We introduce the brat rapid annotation tool (BRAT), an intuitive web-based tool for text annotation supported by Natural Language Processing (NLP) technology. BRAT has been developed for rich structured annotation for a variety of NLP tasks and aims to support manual curation efforts and increase annotator productivity using NLP techniques. We discuss(More)
We introduce a novel compositional language model that works on PredicateArgument Structures (PASs). Our model jointly learns word representations and their composition functions using bagof-words and dependency-based contexts. Unlike previous word-sequencebased models, our PAS-based model composes arguments into predicates by using the category information(More)
We present a novel learning method for word embeddings designed for relation classification. Our word embeddings are trained by predicting words between noun pairs using lexical relation-specific features on a large unlabeled corpus. This allows us to explicitly incorporate relationspecific information into the word embeddings. The learned word embeddings(More)
In this work we propose a novel attentionbased neural network model for the task of fine-grained entity type classification that unlike previously proposed models recursively composes representations of entity mention contexts. Our model achieves state-of-theart performance with 74.94% loose micro F1score on the well-established FIGER dataset, a relative(More)
This paper describes the supporting resources provided for the BioNLP Shared Task 2011. These resources were constructed with the goal to alleviate some of the burden of system development from the participants and allow them to focus on the novel aspects of constructing their event extraction systems. With the availability of these resources we also seek(More)
In this work, we present a novel neural network based architecture for inducing compositional crosslingual word representations. Unlike previously proposed methods, our method fulfills the following three criteria; it constrains the wordlevel representations to be compositional, it is capable of leveraging both bilingual and monolingual data, and it is(More)
In this work, we investigate several neural network architectures for fine-grained entity type classification. Particularly, we consider extensions to a recently proposed attentive neural architecture and make three key contributions. Previous work on attentive neural architectures do not consider hand-crafted features, we combine learnt and hand-crafted(More)
We study two approaches to the marking of extra-propositional aspects of statements in text: the task-independent cue-and-scope representation considered in the CoNLL-2010 Shared Task, and the tagged-event representation applied in several recent event extraction tasks. Building on shared task resources and the analyses from state-of-the-art systems(More)
In this work, we introduce a convolutional neural network model, ConvE, for the task of link prediction. ConvE applies 2D convolution directly on embeddings, thus inducing spatial structure in embedding space. To scale to large knowledge graphs and prevent overfitting due to over-parametrization, previous work seeks to reduce parameters by performing simple(More)