LC-QuAD 2.0: A Large Dataset for Complex Question Answering over Wikidata and DBpedia

Event-QA

“…Often, event-centric information spread across entity-centric knowledge graphs is less annotated and more complex than the entity-centric information and is more difficult to retrieve. As a consequence, existing QA datasets such as LC-QuAD [3,18] and QALD are mainly entity-centric. Specialized event-centric QA datasets are currently non-existing.…”

“…The pipeline starts with a random selection of the type for the semantic query to be generated. The query types included in the Event-QA dataset are the SPARQL query forms ASK and SELECT 3 , as well as the aggregation operator COUNT. ASK queries determine whether a query pattern has a solution.…”

“…Semantic Question Answering (QA) is a key technology to facilitate intuitive access to knowledge graphs for end-users via natural language (NL) interfaces. QA systems automatically translate a user question expressed in a natural language into the SPARQL query language to query knowledge graphs [3,18]. In recent years, researchers have developed a large variety of QA approaches [7].…”

“…QA datasets play an essential role in the development and evaluation of QA systems [3,19]. The active development of QA datasets supports the advancement of QA technology, for example, through the QALD (Question Answering over Linked Data) initiative 1 .…”

“…As popular knowledge graphs are mostly entity-centric, they do not sufficiently cover events and temporal relations [5]. As a consequence, existing QA datasets, with recent examples including LC-QuAD [18], LC-QuAD 2.0 [3] and the QALD challenges, mainly focus on entity-centric queries. More recently, event-centric knowledge graphs such as EventKG [5] that includes historical and important contemporary events as well as knowledge graphs that contain daily happenings extracted from the news (e.g., [9,14]) evolved.…”