Other formats:
BibTeX
LaTeX
RIS
@inproceedings{1876037,
author = {Geletka, Martin and Kalivoda, Vojtěch and Štefánik, Michal and Toma, Marek and Sojka, Petr},
address = {Bologna},
booktitle = {Proceedings of the Working Notes of CLEF 2022 - Conference and Labs of the Evaluation Forum},
editor = {Guglielmo Faggioli, Nicola Ferro, Allan Hanbury, Martin Potthast},
keywords = {information retrieval; question answering; math representations; math-aware information retrieval; word embeddings; ensembling; voting; reranking; data fusion; diversity; transformers},
howpublished = {elektronická verze "online"},
language = {eng},
location = {Bologna},
pages = {28-39},
publisher = {CEUR.org},
title = {Diverse Semantics Representation is King},
url = {http://ceur-ws.org/Vol-3180/paper-02.pdf},
year = {2022}
}
TY - JOUR
ID - 1876037
AU - Geletka, Martin - Kalivoda, Vojtěch - Štefánik, Michal - Toma, Marek - Sojka, Petr
PY - 2022
TI - Diverse Semantics Representation is King
PB - CEUR.org
CY - Bologna
KW - information retrieval
KW - question answering
KW - math representations
KW - math-aware information retrieval
KW - word embeddings
KW - ensembling
KW - voting
KW - reranking
KW - data fusion
KW - diversity
KW - transformers
UR - http://ceur-ws.org/Vol-3180/paper-02.pdf
N2 - We report on the systems that the Math Information Retrieval group at Masaryk University (MIRMU) and the team of Faculty of Informatics students (MSM) prepared for task 1 (find answers) of the ARQMath lab at the CLEF conference. To study the effects of different system settings and hyperparameters, we have prototyped several diverse math-aware information retrieval (MIR) systems: both “old” inverted index-based ones and new neural ones. By ensembling the results of the “weak” individual systems into committees, we report on entailments, benefits, and drawbacks of system ensembling. We evaluated the proposed individual systems and ensembles, considering their diversity, hyperparameters, and representations used, and classified their approaches. Our prototypes have helped to understand the challenging problems of question-answering in the stem domain: the key lies in the proper representation of document semantics. Our reproducible evaluation Python library PV211-utils allows to reproduce and further advance MIR re-search.
ER -
GELETKA, Martin, Vojtěch KALIVODA, Michal ŠTEFÁNIK, Marek TOMA and Petr SOJKA. Diverse Semantics Representation is King. Online. In Guglielmo Faggioli, Nicola Ferro, Allan Hanbury, Martin Potthast. \textit{Proceedings of the Working Notes of CLEF 2022 - Conference and Labs of the Evaluation Forum}. Bologna: CEUR.org, 2022, p.~28-39. ISSN~1613-0073.
|
