The SciQA Scientific Question Answering Benchmark for Scholarly Knowledge.

Knowledge graphs have gained increasing popularity in the last decade in science and technology. However, knowledge graphs are currently relatively simple to moderate semantic structures that are mainly a collection of factual statements. Question answering (QA) benchmarks and systems were so far mainly geared towards encyclopedic knowledge graphs such as DBpedia and Wikidata. We present SciQA a scientific QA benchmark for scholarly knowledge. The benchmark leverages the Open Research Knowledge Graph (ORKG) which includes almost 170,000 resources describing research contributions of almost 15,000 scholarly articles from 709 research fields. Following a bottom-up methodology, we first manually developed a set of 100 complex questions that can be answered using this knowledge graph. Furthermore, we devised eight question templates with which we automatically generated further 2465 questions, that can also be answered with the ORKG. The questions cover a range of research fields and question types and are translated into corresponding SPARQL queries over the ORKG. Based on two preliminary evaluations, we show that the resulting SciQA benchmark represents a challenging task for next-generation QA systems. This task is part of the open competitions at the 22nd International Semantic Web Conference 2023 as the Scholarly Question Answering over Linked Data (QALD) Challenge.

Information extraction pipelines for knowledge graphs.

In the last decade, a large number of knowledge graph (KG) completion approaches were proposed. Albeit effective, these efforts are disjoint, and their collective strengths and weaknesses in effective KG completion have not been studied in the literature. We extend Plumber, a framework that brings together the research community's disjoint efforts on KG completion. We include more components into the architecture of Plumber  to comprise 40 reusable components for various KG completion subtasks, such as coreference resolution, entity linking, and relation extraction. Using these components, Plumber dynamically generates suitable knowledge extraction pipelines and offers overall 432 distinct pipelines. We study the optimization problem of choosing optimal pipelines based on input sentences. To do so, we train a transformer-based classification model that extracts contextual embeddings from the input and finds an appropriate pipeline. We study the efficacy of Plumber for extracting the KG triples using standard datasets over three KGs: DBpedia, Wikidata, and Open Research Knowledge Graph. Our results demonstrate the effectiveness of Plumber in dynamically generating KG completion pipelines, outperforming all baselines agnostic of the underlying KG. Furthermore, we provide an analysis of collective failure cases, study the similarities and synergies among integrated components and discuss their limitations.