Answering Count Questions with Structured Answers from Text

Ghosh, Shrestha; Razniewski, Simon; Weikum, Gerhard

doi:10.1016/j.websem.2022.100769

Cited by 4 publications

(2 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Even so, structured answers are not the norm for more complex or less popular questions. This low recall has prompted research on answering entity counts from multiple SE snippets [6,7] and large-scale mining of quantities from the web [5]. LMs have been shown to be effective in recalling factual information [13,17].…”

Section: Sources Of Informationmentioning

confidence: 99%

“…This paper focuses on dominance estimation: which of two classes has the higher cardinality. We obtain cues for the numeric cardinalities from three sources: the Wikidata KB via SPARQL queries [24], the Bing search engine with judicious queries using the CoQEx method [7], and the GPT-3 language model [2] with various prompts. Absolute cardinalities from these sources are often completely wrong; so we interpret them merely as signals to be used for further inference.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Class Cardinality Comparison as a Fermi Problem

Ghosh¹,

Razniewski²,

Weikum³

2023

Preprint

View full text Add to dashboard Cite

Questions on class cardinality comparisons are quite tricky to answer and come with its own challenges. They require some kind of reasoning since web documents and knowledge bases, indispensable sources of information, rarely store direct answers to questions, such as, "Are there more astronauts or Physics Nobel Laureates?" We tackle questions on class cardinality comparison by tapping into three sources for absolute cardinalities as well as the cardinalities of orthogonal subgroups of the classes. We propose novel techniques for aggregating signals with partial coverage for more reliable estimates and evaluate them on a dataset of 4005 class pairs, achieving an accuracy of 83.7%.

show abstract