Document-level Relation Extraction

The Re-DocRED Dataset resolved the following problems of DocRED: 1. Resolved the incompleteness problem by supplementing large amounts of relation triples. 2. Addressed the logical inconsistencies in DocRED. 3. Corrected the coreferential errors within DocRED.

The 'Deutsche Welle corpus for Information Extraction' (DWIE) is a multi-task dataset that combines four main Information Extraction (IE) annotation sub-tasks: (i) Named Entity Recognition (NER), (ii) Coreference Resolution, (iii) Relation Extraction (RE), and (iv) Entity Linking. DWIE is conceived as an entity-centric dataset that describes interactions and properties of conceptual entities on the level of the complete document. Source: https://arxiv.org/abs/2009.12626

ChemDisGene, a new dataset for training and evaluating multi-class multi-label biomedical relation extraction models.

Bc8BioRED is built upon BioRED 2022 with the addition of directionality annotations. The training and development sets from the original 2022 BioRED corpus were combined and reused as the training set, while the test set was used as the development set. Furthermore, the 400 test abstracts from the BioCreative VIII were utilized for evaluation. Bc8BioRED encompasses seven types of entities and eight types of relationships. Each relationship annotation in the Bc8BioRED corpus is categorized by novelty to indicate whether the relationship represents a significant finding or previously known background knowledge. Initially, the BioRED 2022 corpus comprised 600 abstracts for RE system development, with an additional 400 abstracts annotated to enhance coverage of emerging topics. The dataset encompasses directionality annotations (subject/object roles) for each relation pair, resulting in 10,864 directionality annotations.

The DocRED Information Extraction (DocRED-IE) dataset extends the DocRED dataset for the Document-level Closed Information Extraction (DocIE) task. DocRED-IE is a multi-task dataset and allows for 5 subtasks: (i) Document-level Relation Extraction, (ii) Mention Detection, (iii) Entity Typing, (iv) Entity Disambiguation, (v) Coreference Resolution, as well as combinations thereof such as Named Entity Recognition (NER) or Entity Linking. The DocRED-IE dataset also allows for the end-to-end tasks of: (i) DocIE and (ii) Joint Entity and Relation Extraction. DocRED-IE comprises sentence-level and document-level facts, thereby describing short as well as long-range interactions within an entire document.