Approche de g\'en\'eration de r\'eponse \`a base de transformers (Transformer based approach for answer generation)
Cet article pr{\'e}sente une approche non-supervis{\'e}e bas{\'e}e sur les mod{\e}les Transformer pour la g{\'e}n{\'e}ration du langage naturel dans le cadre des syst{\e}mes de question-r{\'e}ponse. Cette approche permettrait de rem{\'e}dier {\a} la probl{\'e}matique de g{\'e}n{\'e}ration de r{\'e}ponse trop courte ou trop longue sans avoir recours {\a} des donn{\'e}es annot{\'e}es. Cette approche montre des r{\'e}sultats prometteurs pour l{'}anglais et le fran{\c{c}}ais.
Code (0)
등록된 구현이 없습니다.
Tasks
Answer GenerationMethods 이 논문이 사용한 방법론
Linear Layer
A Linear Layer is a projection $\mathbf{XW + b}$.
Absolute Position Encodings
Absolute Position Encodings are a type of position embeddings for [Transformer-based models] where positional encodings are…
Position-Wise Feed-Forward Layer
설명 없음
BPE
Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
Adam
설명 없음
Dropout
Dropout is a regularization technique for neural networks that drops a unit (along with connections) at training time with a specified probability $p$ (a common value is…
Label Smoothing
Label Smoothing is a regularization technique that introduces noise for the labels. This accounts for the fact that datasets may have mistakes in them, so maximizing the…
Multi-Head Attention
설명 없음
Similar Papers 제목 키워드 기반
Ordonnancement de r\'eponses dans les syst\`emes de dialogue bas\'e sur une similarit\'e contexte/r\'eponse (Response ranking in dialogue systems based on context-response similarity)
2018-05-01 · JEPTALNRECITAL 2018 5
· Basma El Amel Boussaha, Hern, Nicolas ez, Christine Jacquin 외
Construire des syst{\`e}mes de dialogue qui conversent avec les humains afin de les aider dans leurs t{\^a}ches quotidiennes est devenu une priorit{\'e}. Certains de ces syst{\`e}mes produisent des dialogues en cherchant…
Typologie des questions \`a r\'eponses multiples pour un syst\`eme de question-r\'eponse (Typology of Multiple Answer Questions for a Question-answering System) [in French]
2012-06-01 · JEPTALNRECITAL 2012 6
· Mathieu-Henri Falco
Question Answering
CELLO: Understanding the responses of data to requests (CELLO : comprendre les r\'eponses des donn\'ees aux requ\^etes) [in French]
2014-07-01 · JEPTALNRECITAL 2014 7
· Yannick Chudy, Yann Desalle, Bruno Gaume, Emmanuel Navarro 외
Web pages segmentation for document selection in Question Answering (Pr\'e-segmentation de pages web et s\'election de documents pertinents en Questions-R\'eponses) [in French]
2013-06-01 · JEPTALNRECITAL 2013 6
· Nicolas Foucault, Sophie Rosset, Gilles Adda
Question Answering
User evaluation of a multiple answer extraction system on the Web (\'Evaluation d'un syst\`eme d'extraction de r\'eponses multiples sur le Web par comparaison \`a des humains) [in French]
2014-07-01 · JEPTALNRECITAL 2014 7
· Mathieu-Henri Falco, V{\'e}ronique Moriceau, Anne Vilnat
Question Answering