Tim Rocktäschel

Learn More
While most approaches to automatically recognizing entailment relations have used classifiers employing hand engineered features derived from complex natural language processing pipelines, in practice their performance has been only slightly better than bag-of-word pair classifiers using only lexical similarity. The only attempt so far to build an(More)
MOTIVATION The accurate identification of chemicals in text is important for many applications, including computer-assisted reconstruction of metabolic networks or retrieval of information about substances in drug development. But due to the diversity of naming conventions and traditions for such molecules, this task is highly complex and should be(More)
Named entity recognition (NER) systems are often based on machine learning techniques to reduce the labor-intensive development of hand-crafted extraction rules and domain-dependent dictionaries. Nevertheless, time-consuming feature engineering is often needed to achieve state-of-the-art performance. In this study, we investigate the impact of such(More)
There are families of neural networks that can learn to compute any function, provided sufficient training data. However, given that in practice training data is scarce for all but a small set of problems, a core question is how to incorporate prior knowledge into a model. Here we consider the case of prior procedural knowledge, such as knowing the overall(More)
Matrix factorization approaches to relation extraction provide several attractive features: they support distant supervision, handle open schemas, and leverage unlabeled data. Unfortunately, these methods share a shortcoming with all other distantly supervised approaches: they cannot learn to extract target relations without existing data in the knowledge(More)
This work describes the participation of the WBI-DDI team on the SemEval 2013 – Task 9.2 DDI extraction challenge. The task consisted of extracting interactions between pairs of drugs from two collections of documents (DrugBank and MEDLINE) and their classification into four subtypes: advise, effect, mechanism, and int. We developed a two-step approach in(More)
Stance detection is the task of classifying the attitude Previous work has assumed that either the target is mentioned in the text or that training data for every target is given. This paper considers the more challenging version of this task, where targets are not always mentioned and no training data is available for the test targets. We experiment with(More)
The automatic extraction of chemical information from text requires the recognition of chemical entity mentions as one of its key steps. When developing supervised named entity recognition (NER) systems, the availability of a large, manually annotated text corpus is desirable. Furthermore, large corpora permit the robust evaluation and comparison of(More)
The BioCreative IV CHEMDNER Task provides participants with the opportunity to compare their methods for chemical named entity recognition (NER) and indexing in a controlled environment. We contributed to this task with our previous conditional random field based system [1] extended by a number of novel general and domain-specific features. For the latter,(More)
Matrix factorization of knowledge bases in universal schema has facilitated accurate distantlysupervised relation extraction. This factorization encodes dependencies between textual patterns and structured relations using lowdimensional vectors defined for each entity pair; although these factors are effective at combining evidence for an entity pair, they(More)