@inproceedings{ri-etal-2022-finding,
title = "Finding Sub-task Structure with Natural Language Instruction",
author = "Ri, Ryokan and
Hou, Yufang and
Marinescu, Radu and
Kishimoto, Akihiro",
editor = "Andreas, Jacob and
Narasimhan, Karthik and
Nematzadeh, Aida",
booktitle = "Proceedings of the First Workshop on Learning with Natural Language Supervision",
month = may,
year = "2022",
address = "Dublin, Ireland",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.lnls-1.1",
doi = "10.18653/v1/2022.lnls-1.1",
pages = "1--9",
abstract = "When mapping a natural language instruction to a sequence of actions, it is often useful toidentify sub-tasks in the instruction. Such sub-task segmentation, however, is not necessarily provided in the training data. We present the A2LCTC (Action-to-Language Connectionist Temporal Classification) algorithm to automatically discover a sub-task segmentation of an action sequence.A2LCTC does not require annotations of correct sub-task segments and learns to find them from pairs of instruction and action sequence in a weakly-supervised manner. We experiment with the ALFRED dataset and show that A2LCTC accurately finds the sub-task structures. With the discovered sub-tasks segments, we also train agents that work on the downstream task and empirically show that our algorithm improves the performance.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ri-etal-2022-finding">
<titleInfo>
<title>Finding Sub-task Structure with Natural Language Instruction</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ryokan</namePart>
<namePart type="family">Ri</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yufang</namePart>
<namePart type="family">Hou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Radu</namePart>
<namePart type="family">Marinescu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Akihiro</namePart>
<namePart type="family">Kishimoto</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2022-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the First Workshop on Learning with Natural Language Supervision</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jacob</namePart>
<namePart type="family">Andreas</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Karthik</namePart>
<namePart type="family">Narasimhan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Aida</namePart>
<namePart type="family">Nematzadeh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Dublin, Ireland</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>When mapping a natural language instruction to a sequence of actions, it is often useful toidentify sub-tasks in the instruction. Such sub-task segmentation, however, is not necessarily provided in the training data. We present the A2LCTC (Action-to-Language Connectionist Temporal Classification) algorithm to automatically discover a sub-task segmentation of an action sequence.A2LCTC does not require annotations of correct sub-task segments and learns to find them from pairs of instruction and action sequence in a weakly-supervised manner. We experiment with the ALFRED dataset and show that A2LCTC accurately finds the sub-task structures. With the discovered sub-tasks segments, we also train agents that work on the downstream task and empirically show that our algorithm improves the performance.</abstract>
<identifier type="citekey">ri-etal-2022-finding</identifier>
<identifier type="doi">10.18653/v1/2022.lnls-1.1</identifier>
<location>
<url>https://aclanthology.org/2022.lnls-1.1</url>
</location>
<part>
<date>2022-05</date>
<extent unit="page">
<start>1</start>
<end>9</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Finding Sub-task Structure with Natural Language Instruction
%A Ri, Ryokan
%A Hou, Yufang
%A Marinescu, Radu
%A Kishimoto, Akihiro
%Y Andreas, Jacob
%Y Narasimhan, Karthik
%Y Nematzadeh, Aida
%S Proceedings of the First Workshop on Learning with Natural Language Supervision
%D 2022
%8 May
%I Association for Computational Linguistics
%C Dublin, Ireland
%F ri-etal-2022-finding
%X When mapping a natural language instruction to a sequence of actions, it is often useful toidentify sub-tasks in the instruction. Such sub-task segmentation, however, is not necessarily provided in the training data. We present the A2LCTC (Action-to-Language Connectionist Temporal Classification) algorithm to automatically discover a sub-task segmentation of an action sequence.A2LCTC does not require annotations of correct sub-task segments and learns to find them from pairs of instruction and action sequence in a weakly-supervised manner. We experiment with the ALFRED dataset and show that A2LCTC accurately finds the sub-task structures. With the discovered sub-tasks segments, we also train agents that work on the downstream task and empirically show that our algorithm improves the performance.
%R 10.18653/v1/2022.lnls-1.1
%U https://aclanthology.org/2022.lnls-1.1
%U https://doi.org/10.18653/v1/2022.lnls-1.1
%P 1-9
Markdown (Informal)
[Finding Sub-task Structure with Natural Language Instruction](https://aclanthology.org/2022.lnls-1.1) (Ri et al., LNLS 2022)
ACL