About
InVeRo
InVeRo (Intelligible Verbs and Roles) is a platform created with the aim of making Semantic Role Labeling - the task of answering "Who did What to Whom, When and Where?" - more accessible to a wider audience. Indeed, Semantic Role Labeling presents intrinsic difficulties, from the complexity of traditional language resources used in the task to the level of intricacy of current state-of-the-art systems. In order to achieve its goal:
- InVeRo uses the intelligible verb senses and semantic roles of a recently proposed linguistic resource, VerbAtlas.
- InVeRo offers an off-the-shelf full end-to-end system, InVeRo-SRL, to obtain semantically-annotated sentences without writing a single line of code.
The InVeRo platform makes these tools accessible through a user-friendly Web interface and while also featuring easy-to-use REST APIs to directly obtain VerbAtlas-specific information or annotate sentences with predicate sense and semantic role labels.
VerbAtlas
VerbAtlas is a hand-crafted lexical-semantic resource whose goal is to bring together all verbal synsets from BabelNet into semantically-coherent frames. The frames define a common, prototypical argument structure while at the same time providing new concept-specific information. VerbAtlas comes with an explicit, cross-frame set of semantic roles linked to selectional preferences expressed in terms of BabelNet synsets, and is the first resource enriched with semantic information about implicit, shadow, and default arguments.
This website provides:
- a web interface for looking up terms and Prototypical Argument Structures (with hyperlinks pointing to BabelNet);
- statistics, a bibliography and a download section.
References
Andrea di Fabio, Simone Conia, Roberto Navigli. VerbAtlas: a Novel Large-Scale Verbal Semantic Resource and Its Application to Semantic Role Labeling. Proc. of the 2019 Conference on Empirical Methods in Natural Language Processing (EMNLP 2019), Hong Kong, China, November 3-7, 2019.
The authors gratefully acknowledge the support of the ERC Consolidator Grant MOUSSE No. 726487 and the ELEXIS project No. 731015 under the European Union’s Horizon 2020 research and innovation programme.
InVeRo-SRL
InVeRo-SRL, part of the InVeRo platform, is an off-the-shelf state-of-the-art system for span-based Semantic Role Labeling with PropBank and the recently proposed VerbAtlas.
Our system tackles the traditional Semantic Role Labeling pipeline in each of its parts, from predicate identification to predicate sense disambiguation, from argument identification to argument classification.
InVeRo-SRL also lowers the requirements for Semantic Role Labeling annotations as it does not require any syntactic information at the input-level, that is, a user just has to provide a raw text sentence to the system to get its corresponding predicate sense and semantic role annotations.
This project is funded by the MOUSSE ERC Grant no.726487 and the ELEXIS project no.731015 under the European Union's Horizon 2020 research and innovation programme.
VerbAtlas is licensed under the CC BY-NC-SA 4.0 License