Structured sentiment analysis attempts to extract full opinion tuples from a text, but over time this task has been subdivided into smaller and smaller sub-tasks, e.g., target extraction or targeted polarity classification. We argue that this division has become counterproductive and propose a new unified framework to remedy the situation. We cast the structured sentiment problem as dependency graph parsing, where the nodes are spans of sentiment holders, targets and expressions, and the arcs are the relations between them. We perform experiments on five datasets in four languages (English, Norwegian, Basque, and Catalan) and show that this approach leads to strong improvements over state-of-the-art baselines. Our analysis shows that refining the sentiment graphs with syntactic dependency information further improves results.

We present a neural end-to-end architecture for negation resolution based on a formulation of the task as a graph parsing problem. Our approach allows for the straightforward inclusion of many types of graph-structured features without the need for representationspecific heuristics. In our experiments, we specifically gauge the usefulness of syntactic information for negation resolution. Despite the conceptual simplicity of our architecture, we achieve state-of-the-art results on the Conan Doyle benchmark dataset, including a new top result for our best model.

No abstract

The development of accurate machine translation systems requires detailed analyses of the recurring translation mistakes. However, the manual inspection of the decoder log files is a daunting task because of their sheer size and their uncomfortable format, in which the relevant data is widely spread. For all major platforms, DIMwid offers a graphical user interface that allows the quick inspection of the decoder stacks or chart cells for a given span in a uniform way. Currently, DIMwid can process the decoder log files of the phrase-based stack decoder and the syntax-based chart decoder inside the Moses framework.

No abstract

This paper demonstrates how a graph-based semantic parser can be applied to the task of structured sentiment analysis, directly predicting sentiment graphs from text. We advance the state of the art on 4 out of 5 standard benchmark sets. We release the source code, models and predictions. 1

This paper demonstrates how a graph-based semantic parser can be applied to the task of structured sentiment analysis, directly predicting sentiment graphs from text. We advance the state of the art on 4 out of 5 standard benchmark sets. We release the source code, models and predictions. 1

This article provides an account of the making of KBLab, the data lab at the National Library of Sweden (KB). The first part of the article offers an evaluative discussion of the work involved in establishing KBLab as both a physical and a digital site for researchers to use KB’s digital collections at previously unimaginable scales. Beyond explaining how the lab aligns with KB’s broader mission as a national library, we also elaborate upon the design of the technical setup and the processes of research coordination that the operation of a library lab presumes. The second part discusses how KBLab has deployed the library’s collections as data to produce high quality Swedish AI models, which constitute a significant new form of digital research infrastructure. We situate this development work in the context of uneven AI coverage for smaller languages, and consider how the lab’s models have contributed to the making of important AI infrastructure for the Swedish language. The conclusion raises the possibilities and challenges involved in continuing the type of library-based AI development we have initiated at KBLab.