Exploration of neural machine translation in autoformalization of mathematics in Mizar

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21730%2F20%3A00346154" target="_blank" >RIV/68407700:21730/20:00346154 - isvavai.cz</a>

Výsledek na webu

<a href="https://doi.org/10.1145/3372885.3373827" target="_blank" >https://doi.org/10.1145/3372885.3373827</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1145/3372885.3373827" target="_blank" >10.1145/3372885.3373827</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Exploration of neural machine translation in autoformalization of mathematics in Mizar

Popis výsledku v původním jazyce

In this paper we share several experiments trying to automatically translate informal mathematics into formal mathematics. In our context informal mathematics refers to humanwritten mathematical sentences in the LaTeX format; and formal mathematics refers to statements in the Mizar language. We conducted our experiments against three established neural network-based machine translation models that are known to deliver competitive results on translating between natural languages. To train these models we also prepared four informal-to-formal datasets. We compare and analyze our results according to whether the model is supervised or unsupervised. In order to augment the data available for auto-formalization and improve the results, we develop a custom type-elaboration mechanism and integrate it in the supervised translation.

Název v anglickém jazyce

Exploration of neural machine translation in autoformalization of mathematics in Mizar

Popis výsledku anglicky

In this paper we share several experiments trying to automatically translate informal mathematics into formal mathematics. In our context informal mathematics refers to humanwritten mathematical sentences in the LaTeX format; and formal mathematics refers to statements in the Mizar language. We conducted our experiments against three established neural network-based machine translation models that are known to deliver competitive results on translating between natural languages. To train these models we also prepared four informal-to-formal datasets. We compare and analyze our results according to whether the model is supervised or unsupervised. In order to augment the data available for auto-formalization and improve the results, we develop a custom type-elaboration mechanism and integrate it in the supervised translation.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/EF15_003%2F0000466" target="_blank" >EF15_003/0000466: Umělá inteligence a uvažování</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2020

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

CPP 2020: Proceedings of the 9th ACM SIGPLAN International Conference on Certified Programs and Proofs

ISBN

978-1-4503-7097-4

ISSN

—

e-ISSN

—

Počet stran výsledku

14

Strana od-do

85-98

Název nakladatele

Association for Computing Machinery

Místo vydání

New York

Místo konání akce

New Orleans

Datum konání akce

20. 1. 2020

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000620147900009