Sentiment analysis can be regarded as a relation extraction problem in which the sentiment of some opinion holder towards a certain aspect of a product, theme or event needs to be extracted. We present a novel neural architecture for sentiment analysis as a relation extraction problem that addresses this problem by dividing it into three subtasks: i) identification of aspect and opinion terms, ii) labeling of opinion terms with a sentiment, and iii) extraction of relations between opinion terms and aspect terms. For each subtask, we propose a neural network based component and combine all of them into a complete system for relational sentiment analysis.The component for aspect and opinion term extraction is a hybrid architecture consisting of a recurrent neural network stacked on top of a convolutional neural network. This approach outperforms a standard convolutional deep neural architecture as well as a recurrent network architecture and performs competitively compared to other methods on two datasets of annotated customer reviews. To extract sentiments for individual opinion terms, we propose a recurrent architecture in combination with word distance features and achieve promising results, outperforming a majority baseline by 18% accuracy and providing the first results for the USAGE dataset. Our relation extraction component outperforms the current state-of-the-art in aspect-opinion relation extraction by 15% F-Measure.

Word embeddings have been shown to be highly effective in a variety of lexical semantic tasks. They tend to capture meaningful relational similarities between individual words, at the expense of lacking the capabilty of making the underlying semantic relation explicit. In this paper, we investigate the attribute relation that often holds between the constituents of adjective-noun phrases. We use CBOW word embeddings to represent word meaning and learn a compositionality function that combines the individual constituents into a phrase representation, thus capturing the compositional attribute meaning. The resulting embedding model, while being fully interpretable, outperforms countbased distributional vector space models that are tailored to attribute meaning in the two tasks of attribute selection and phrase similarity prediction. Moreover, as the model captures a generalized layer of attribute meaning, it bears the potential to be used for predictions over various attribute inventories without re-training.

International audienceThe paper presents an approach to extract knowledge from large text corpora, in particular knowledge that facilitates object manipulation by embodied intelligent systems that need to act in the world. As a first step, our goal is to extract the prototypical location of given objects from text corpora. We approach this task by calculating relatedness scores for objects and locations using techniques from distributional semantics. We empirically compare different methods for representing locations and objects as vectors in some geometric space, and we evaluate them with respect to a crowd-sourced gold standard in which human subjects had to rate the prototypicality of a location given an object. By applying the proposed framework on DBpedia, we are able to build a knowledge base of 931 high confidence object-locations relations in a fully automatic fashion.

The task of answering natural language questions over RDF data has received wide interest in recent years, in particular in the context of the series of QALD benchmarks. The task consists of mapping a natural language question to an executable form, e.g. SPARQL, so that answers from a given KB can be extracted. So far, most systems proposed are i) monolingual and ii) rely on a set of hard-coded rules to interpret questions and map them into a SPARQL query. We present the first multilingual QALD pipeline that induces a model from training data for mapping a natural language question into logical form as probabilistic inference. In particular, our approach learns to map universal syntactic dependency representations to a language-independent logical form based on DUDES (Dependency-based Underspecified Discourse Representation Structures) that are then mapped to a SPARQL query as a deterministic second step. Our model builds on factor graphs that rely on features extracted from the dependency graph and corresponding semantic representations. We rely on approximate inference techniques, Markov Chain Monte Carlo methods in particular, as well as Sample Rank to update parameters using a ranking objective. Our focus lies on developing methods that overcome the lexical gap and present a novel combination of machine translation and word embedding approaches for this purpose. As a proof of concept for our approach, we evaluate our approach on the QALD-6 datasets for English, German & Spanish.

Abstract. Named Entity Disambiguation (NED) is the task of disambiguating named entities in a natural language text by linking them to their corresponding entities in a knowledge base such as DBpedia, which are already recognized. It is an important step in transforming unstructured text into structured knowledge. Previous work on this task has proven a strong impact of graph-based methods such as PageRank on entity disambiguation. Other approaches rely on distributional similarity between an article and the textual description of a candidate entity. However, the combined impact of these different feature groups has not been explored to a sufficient extent. In this paper, we present a novel approach that exploits an undirected probabilistic model to combine different types of features for named entity disambiguation. Capitalizing on Markov Chain Monte Carlo sampling, our model is capable of exploiting complementary strengths between both graph-based and textual features. We analyze the impact of these features and their combination on named entity disambiguation. In an evaluation on the GERBIL benchmark, our model compares favourably to the current state-of-the-art in 8 out of 14 data sets.