SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2004.07347 · EMNLP Findings · 2020

HybridQA: A Dataset of Multi-Hop Question Answering over Tabular and Textual Data

Zhiyu Chen, Wenhu Chen, William Wang, Hong Wang, Hanwen Zha, Wenhan Xiong

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 15 functions out of this paper's own repositories and ran 12 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
wenhuchen/HybridQA canonical 3 of 3
teacherpeterpan/Unsupervised-Multi-hop-QA pwc_unofficial 7 of 10
copy not recorded — 2 of 2
FunctionStatusWhere it lives
compute_exact Ran wenhuchen/HybridQA/evaluate_script.py
code served (permissive licence) · get_code("9026d9da5d085bd4")
get_bridge_entities Ran teacherpeterpan/Unsupervised-Multi-hop-QA/MQA_QG/HybridQA/operations.py
code served (permissive licence) · get_code("a745217943d329bd")
get_tokens Ran this paper's copy was not recorded; identical code first harvested from saptarshi059/cdqa-project
pointer only · get_code("aa6bc32d9c17a2f1")
include_bridge_entity Ran teacherpeterpan/Unsupervised-Multi-hop-QA/MQA_QG/HotpotQA/operations.py
code served (permissive licence) · get_code("c7994a1f51a22564")
isNaN Ran wenhuchen/HybridQA/train_stage12.py
code served (permissive licence) · get_code("86df5eac4985fe0e")
load_all_questions Ran teacherpeterpan/Unsupervised-Multi-hop-QA/MQA_QG/run_ppl_ranking.py
code served (permissive licence) · get_code("199bd77e59857d70")
load_hotpotQA Ran teacherpeterpan/Unsupervised-Multi-hop-QA/prep_data_hotpotQA.py
code served (permissive licence) · get_code("b8a1108847521ccb")
normalize_answer Ran this paper's copy was not recorded; identical code first harvested from saptarshi059/cdqa-project
pointer only · get_code("6a96435eba311b08")
phrase_match Ran teacherpeterpan/Unsupervised-Multi-hop-QA/MQA_QG/HotpotQA/operations.py
code served (permissive licence) · get_code("fd256c34b37a0e87")
phrase_overlap Ran teacherpeterpan/Unsupervised-Multi-hop-QA/MQA_QG/HybridQA/utils.py
code served (permissive licence) · get_code("8a1e9b4aa699789a")
title_match Ran teacherpeterpan/Unsupervised-Multi-hop-QA/prep_data_hotpotQA.py
code served (permissive licence) · get_code("8f0d111c90b80c13")
url2text Ran wenhuchen/HybridQA/train_stage12.py
code served (permissive licence) · get_code("38c3f4fa8e089762")
construct_sample Not yet run teacherpeterpan/Unsupervised-Multi-hop-QA/MQA_QG/run_multihop_generation.py
code served (permissive licence) · get_code("f1a4954ff85e0b97")
convert_question_into_desc Not yet run teacherpeterpan/Unsupervised-Multi-hop-QA/MQA_QG/HybridQA/operations.py
code served (permissive licence) · get_code("79bc6415d0faf9c6")
get_passage Not yet run teacherpeterpan/Unsupervised-Multi-hop-QA/MQA_QG/HybridQA/operations.py
code served (permissive licence) · get_code("04a978d51168a6db")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Existing question answering datasets focus on dealing with homogeneous information, based either only on text or KB/Table information alone. However, as human knowledge is distributed over heterogeneous forms, using homogeneous information alone might lead to severe coverage problems. To fill in the gap, we present HybridQA 1 , a new large-scale question-answering dataset that requires reasoning on heterogeneous information. Each question is aligned with a Wikipedia table and multiple free-form corpora linked with the entities in the table. The questions are designed to aggregate both tabular information and text information, i.e., lack of either form would render the question unanswerable. We test with three different models: 1) a table-only model. 2) text-only model. 3) a hybrid model that combines heterogeneous information to find the answer. The experimental results show that the EM scores obtained by two baselines are below 20%, while the hybrid model can achieve an EM over 40%. This gap suggests the necessity to aggregate heterogeneous information in HybridQA. However, the hybrid model's score is still far behind human performance. Hence, HybridQA can serve as a challenging benchmark to study question answering with heterogeneous information.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2004.07347")
get_code_for_paper("2004.07347")
have("2004.07347")

Connect an agent — have() is free.