Skip to content

Latest commit

 

History

History
152 lines (136 loc) · 20.8 KB

qald.md

File metadata and controls

152 lines (136 loc) · 20.8 KB

Question Answering over Linked Data (QALD)

QALD is a series of evaluation campaigns on question answering over linked data, which aims at providing an up-to-date benchmark for assessing and comparing state-of-the-art systems that mediate between a user, expressing his or her information need in natural language, and RDF data. Thus, it targets all researchers and practitioners working on querying Linked Data, natural language processing for question answering, multilingual information retrieval and related topics. The main goal is to gain insights into the strengths and shortcomings of different approaches and into possible solutions for coping with the large, heterogeneous and distributed nature of Semantic Web data.

QALD challenge began in 2011 and is developing benchmarks that are increasingly being used as standard evaluation venue for question answering over Linked Data. Overviews of past instantiations of the challenge are available from the CLEF Working Notes, CEUR workshop notes as well as ESWC proceedings.

The key challenge for QA over Linked Data is to translate a user's natural language query into such a form that it can be evaluated using standard Semantic Web query processing and inferencing techniques. The main task of QALD therefore is the following:

Given one or several RDF dataset(s) as well as additional knowledge sources and natural language questions or keywords, return the correct answers or a SPARQL query that retrieves these answers.

Table of contents

QALD-10

Please see the original paper for details about the dataset creation process, data format, task and participating systems.

Leaderboard

Model / System Year Precision Recall F1 Language Reported by
Borroto et al. (SPARQL-QA) 2022 45.38 (Macro) 45.74 (Macro) 59.47 (Macro) EN GERBIL
QAnswer 2022 50.68 (Macro) 52.38 (Macro) 57.76 (Macro) EN GERBIL
Steinmetz et al. 2022 32.06 (Macro) 33.12 (Macro) 49.09 (Macro) EN GERBIL
Baramiia et al. 2022 42.89 (Macro) 42.72 (Macro) 42.81 (Macro) EN GERBIL
Gavrilev et al. 2022 14.21 (Macro) 14.00 (Macro) 19.48 (Macro) EN GERBIL

QALD-9-PLUS

Please see the original paper for details about the dataset creation process, data format, task and participating systems.

Leaderboard

Model / System Year Precision Recall F1 Language Reported by
QAnswer 2022 50.46 44.84 44.59 EN Perevalov et. al.
QAnswer-en-de-Yandex 2022 34.91 35.18 32.83 EN Perevalov et. al.
QAnswer-en-de-Helsinki NLP 2022 34.90 35.00 33.74 EN Perevalov et. al.
QAnswer-en-ru-Yandex 2022 22.84 21.61 21.07 EN Perevalov et. al.
QAnswer-en-ru-Helsinki NLP 2022 22.38 20.63 20.41 EN Perevalov et. al.
QAnswer-en-fr-Yandex 2022 32.16 29.49 28.86 EN Perevalov et. al.
QAnswer-en-fr-Helsinki NLP 2022 30.76 28.19 27.57 EN Perevalov et. al.
QAnswer 2022 33.05 32.44 31.71 DE Perevalov et. al.
QAnswer-de-en-Yandex 2022 41.36 38.71 36.49 DE Perevalov et. al.
QAnswer-de-en-Helsinki NLP 2022 40.95 38.11 36.57 DE Perevalov et. al.
QAnswer-de-ru-Yandex 2022 22.33 21.61 21.07 DE Perevalov et. al.
QAnswer-de-fr-Yandex 2022 31.05 28.23 27.71 DE Perevalov et. al.
QAnswer-de-fr-Helsinki NLP 2022 28.32 26.21 25.21 DE Perevalov et. al.
QAnswer 2022 22.29 22.23 21.43 RU Perevalov et. al.
QAnswer-ru-en-Yandex 2022 44.09 39.79 38.98 RU Perevalov et. al.
QAnswer-ru-en-Helsinki NLP 2022 39.12 36.02 34.63 RU Perevalov et. al.
QAnswer-ru-de-Yandex 2022 29.20 29.88 27.17 RU Perevalov et. al.
QAnswer-ru-fr-Yandex 2022 29.82 27.34 26.62 RU Perevalov et. al.
QAnswer-ru-fr-Helsinki NLP 2022 21.60 20.44 19.72 RU Perevalov et. al.
QAnswer 2022 25.32 25.35 23.00 FR Perevalov et. al.
QAnswer-fr-en-Yandex 2022 42.11 35.04 36.57 FR Perevalov et. al.
QAnswer-fr-en-Helsinki NLP 2022 42.11 35.04 36.57 FR Perevalov et. al.
QAnswer-fr-de-Yandex 2022 33.01 35.82 31.01 FR Perevalov et. al.
QAnswer-fr-de-Helsinki NLP 2022 27.74 30.56 25.74 FR Perevalov et. al.
QAnswer-fr-ru-Yandex 2022 11.58 15.74 12.28 FR Perevalov et. al.
QAnswer-fr-ru-Helsinki NLP 2022 0 0 0 FR Perevalov et. al.
QAnswer-lt-en-Yandex 2022 41.96 39.19 39.30 LT Perevalov et. al.
QAnswer-lt-de-Yandex 2022 38.94 36.60 37.34 LT Perevalov et. al.
QAnswer-lt-de-Helsinki NLP 2022 45.73 39.35 39.49 LT Perevalov et. al.
QAnswer-lt-ru-Yandex 2022 17.32 17.39 17.35 LT Perevalov et. al.
QAnswer-lt-ru-Helsinki NLP 2022 14.13 15.22 14.49 LT Perevalov et. al.
QAnswer-lt-fr-Yandex 2022 33.26 28.32 29.16 LT Perevalov et. al.
QAnswer-lt-fr-Helsinki NLP 2022 30.36 26.28 26.42 LT Perevalov et. al.
QAnswer-uk-en-Yandex 2022 43.15 43.15 39.17 UK Perevalov et. al.
QAnswer-uk-en-Helsinki NLP 2022 32.67 33.01 30.38 UK Perevalov et. al.
QAnswer-uk-de-Yandex 2022 31.86 30.32 28.82 UK Perevalov et. al.
QAnswer-uk-de-Helsinki NLP 2022 29.48 27.32 27.13 UK Perevalov et. al.
QAnswer-uk-ru-Yandex 2022 22.65 22.53 22.02 UK Perevalov et. al.
QAnswer-uk-ru-Helsinki NLP 2022 24.64 23.97 23.19 UK Perevalov et. al.
QAnswer-uk-fr-Yandex 2022 30.85 28.07 27.27 UK Perevalov et. al.
QAnswer-uk-fr-Helsinki NLP 2022 22.68 23.52 20.76 UK Perevalov et. al.
QAnswer-be-en-Yandex 2022 54.55 45.54 45.62 BE Perevalov et. al.
QAnswer-be-de-Yandex 2022 36.54 27.47 27.46 BE Perevalov et. al.
QAnswer-be-ru-Yandex 2022 18.18 18.18 18.18 BE Perevalov et. al.
QAnswer-be-fr-Yandex 2022 40.91 22.81 24.41 BE Perevalov et. al.
QAnswer-hy-en-Yandex 2022 26.32 21.88 22.53 HY Perevalov et. al.
QAnswer-hy-de-Yandex 2022 27.74 34.95 29.50 HY Perevalov et. al.
QAnswer-hy-ru-Yandex 2022 5.26 5.26 5.26 HY Perevalov et. al.
QAnswer-hy-fr-Yandex 2022 16.84 16.6 13.78 HY Perevalov et. al.
QAnswer-ba-en-Yandex 2022 31.42 28.45 28.44 BA Perevalov et. al.
QAnswer-ba-de-Yandex 2022 23.66 23.19 22.87 BA Perevalov et. al.
QAnswer-ba-ru-Yandex 2022 20.73 20.16 19.17 BA Perevalov et. al.
QAnswer-ba-fr-Yandex 2022 18.63 17.23 16.08 BA Perevalov et. al.
DeepPavlov 2022 17.65 11.30 12.40 EN Perevalov et. al.
DeepPavlov-en-ru-Yandex 2022 7.35 6.99 7.11 EN Perevalov et. al.
DeepPavlov-en-ru-Helsinki NLP 2022 8.82 7.35 7.84 EN Perevalov et. al.
DeepPavlov-de-en-Yandex 2022 19.85 12.59 13.92 DE Perevalov et. al.
DeepPavlov-de-en-Helsinki NLP 2022 16.18 9.83 10.93 DE Perevalov et. al.
DeepPavlov-de-ru-Yandex 2022 8.82 7.37 7.63 DE Perevalov et. al.
DeepPavlov 2022 9.56 8.33 8.70 RU Perevalov et. al.
DeepPavlov-ru-en-Yandex 2022 16.91 10.54 11.61 RU Perevalov et. al.
DeepPavlov-ru-en-Helsinki NLP 2022 13.24 7.93 8.79 RU Perevalov et. al.
DeepPavlov-fr-en-Yandex 2022 31.58 22.02 22.73 FR Perevalov et. al.
DeepPavlov-fr-en-Helsinki NLP 2022 21.05 16.67 17.29 FR Perevalov et. al.
DeepPavlov-fr-ru-Yandex 2022 10.53 10.53 10.53 FR Perevalov et. al.
DeepPavlov-fr-ru-Helsinki NLP 2022 10.53 10.53 10.53 FR Perevalov et. al.
DeepPavlov-lt-en-Yandex 2022 17.39 10.94 11.73 LT Perevalov et. al.
DeepPavlov-lt-ru-Yandex 2022 4.35 4.35 4.35 LT Perevalov et. al.
DeepPavlov-lt-ru-Helsinki NLP 2022 4.35 4.35 4.35 LT Perevalov et. al.
DeepPavlov-uk-en-Yandex 2022 16.91 10.18 11.37 UK Perevalov et. al.
DeepPavlov-uk-en-Helsinki NLP 2022 16.91 10.24 11.18 UK Perevalov et. al.
DeepPavlov-uk-ru-Yandex 2022 7.35 6.99 7.11 UK Perevalov et. al.
DeepPavlov-uk-ru-Helsinki NLP 2022 6.62 6.25 6.37 UK Perevalov et. al.
DeepPavlov-be-en-Yandex 2022 9.09 4.55 6.06 BE Perevalov et. al.
DeepPavlov-be-ru-Yandex 2022 9.09 9.09 9.09 BE Perevalov et. al.
DeepPavlov-hy-en-Yandex 2022 21.05 11.49 12.21 HY Perevalov et. al.
DeepPavlov-hy-ru-Yandex 2022 10.53 10.53 10.53 HY Perevalov et. al.
DeepPavlov-ba-en-Yandex 2022 15.62 9.07 9.86 BA Perevalov et. al.
DeepPavlov-ba-ru-Yandex 2022 9.38 7.55 8.06 BA Perevalov et. al.
Platypus 2022 15.89 15.26 15.03 EN Perevalov et. al.
Platypus-en-fr-Yandex 2022 8.94 10.10 8.92 EN Perevalov et. al.
Platypus-en-fr-Helsinki NLP 2022 8.21 9.36 8.19 EN Perevalov et. al.
Platypus-de-en-Yandex 2022 12.27 11.60 11.44 DE Perevalov et. al.
Platypus-de-en-Helsinki NLP 2022 15.21 14.34 14.26 DE Perevalov et. al.
Platypus-de-fr-Yandex 2022 12.27 11.60 11.44 DE Perevalov et. al.
Platypus-de-fr-Helsinki NLP 2022 5.91 6.62 5.93 DE Perevalov et. al.
Platypus-ru-en-Yandex 2022 16.13 15.27 15.11 RU Perevalov et. al.
Platypus-ru-en-Helsinki NLP 2022 13.24 7.93 8.79 RU Perevalov et. al.
Platypus-ru-fr-Yandex 2022 7.84 9.19 7.81 RU Perevalov et. al.
Platypus-ru-fr-Helsinki NLP 2022 6.64 7.35 6.67 RU Perevalov et. al.
Platypus 2022 4.17 4.17 4.17 FR Perevalov et. al.
Platypus-fr-en-Yandex 2022 21.05 21.05 21.05 FR Perevalov et. al.
Platypus-fr-en-Helsinki NLP 2022 21.05 21.05 21.05 FR Perevalov et. al.
Platypus-lt-en-Yandex 2022 17.39 15.22 15.94 LT Perevalov et. al.
Platypus-lt-fr-Yandex 2022 13.59 15.22 12.56 LT Perevalov et. al.
Platypus-lt-fr-Helsinki NLP 2022 13.04 13.04 13.04 LT Perevalov et. al.
Platypus-uk-en-Yandex 2022 14.97 14.35 14.17 UK Perevalov et. al.
Platypus-uk-en-Helsinki NLP 2022 14.22 13.44 13.48 UK Perevalov et. al.
Platypus-uk-fr-Yandex 2022 9.68 10.66 9.53 UK Perevalov et. al.
Platypus-uk-fr-Helsinki NLP 2022 7.75 8.46 7.65 UK Perevalov et. al.
QAnswer-be-en-Yandex 2022 9.09 9.09 9.09 BE Perevalov et. al.
QAnswer-be-fr-Yandex 2022 9.09 9.09 9.09 BE Perevalov et. al.
QAnswer-hy-en-Yandex 2022 15.79 15.79 15.79 HY Perevalov et. al.
QAnswer-hy-fr-Yandex 2022 5.26 5.26 5.26 HY Perevalov et. al.
QAnswer-ba-en-Yandex 2022 8.77 8.36 7.95 BA Perevalov et. al.
QAnswer-ba-fr-Yandex 2022 5.24 6.25 5.28 BA Perevalov et. al.