Feedback

Faculté des Sciences appliquées
Faculté des Sciences appliquées
MASTER THESIS
VIEW 93 | DOWNLOAD 543

Automatic Abstractive Text Summarization : A deeper look into convolutional sequence-to-sequence networks

Download
Vermeylen, Valentin ULiège
Promotor(s) : Ittoo, Ashwin ULiège ; Doloris, Samy
Date of defense : 6-Sep-2021/7-Sep-2021 • Permalink : http://hdl.handle.net/2268.2/13292
Details
Title : Automatic Abstractive Text Summarization : A deeper look into convolutional sequence-to-sequence networks
Translated title : [fr] Synthétisation Abstractive et Automatique de Textes : Un examen des réseaux séquence-vers-séquence convolutionnels
Author : Vermeylen, Valentin ULiège
Date of defense  : 6-Sep-2021/7-Sep-2021
Advisor(s) : Ittoo, Ashwin ULiège
Doloris, Samy 
Committee's member(s) : Fontaine, Pascal ULiège
Gribomont, Pascal ULiège
Language : English
Number of pages : 65
Keywords : [en] abstractive summarization
[en] convolutional sequence-to-sequence
Discipline(s) : Engineering, computing & technology > Computer science
Funders : NRB
Target public : General public
Institution(s) : Université de Liège, Liège, Belgique
Degree: Master en ingénieur civil en informatique, à finalité spécialisée en "intelligent systems"
Faculty: Master thesis of the Faculté des Sciences appliquées

Abstract

[en] As the amount of information produced everyday continually increases, the desire for summaries containing only the most salient parts of the texts continues to gain traction. Even though the possibility to extract parts of texts and gluing them together already exists, we usually prefer fluent, human-like summaries.

That is the concern of the Artificial Intelligence subfield of Automatic Abstractive Summarization. Although the task is typically solved using recurrent neural networks, that architecture comes with several challenges, the biggest being the amount of time and computational power required to train the models. Fortunately, another less computationally intensive paradigm exists, based on convolutional networks, even though it has not been as extensively studied.

This thesis is concerned with that convolutional framework, and explores questions and assumptions that have not been answered previously, such as the advantages and drawbacks of using pretrained embeddings, or the tradeoff between performance gains and the added complexity of mechanisms such as reinforcement learning or pointing-generation. Experiments about the abstractiveness of the models, their fine-tuning on a different dataset, and their ability to capture long-distanced dependencies are also performed through the use of both the CNN/DailyMail dataset, and the XSUM dataset.

Those experiments show that using more convolutional blocks in the model makes sense up to a certain point, that the use of pretrained embeddings is advisable, as is the use of the pointer-generator network implemented in this work. The use of reinforcement learning is also advisable at the end of the model training.

Finally, this thesis is concluded with additional experiments that could be implemented in future works, as well as practical advises regarding the use of abstractive summarization in the context of general terms and conditions summarization.


File(s)

Document(s)

File
Access Master_Thesis_Vermeylen_Valentin.pdf
Description:
Size: 1.47 MB
Format: Adobe PDF
File
Access Abstract_Vermeylen_Valentin.pdf
Description: Résumé
Size: 191.8 kB
Format: Adobe PDF

Author

  • Vermeylen, Valentin ULiège Université de Liège > Master ingé. civ. info., à fin.

Promotor(s)

Committee's member(s)

  • Fontaine, Pascal ULiège Université de Liège - ULiège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Systèmes informatiques distribués
    ORBi View his publications on ORBi
  • Gribomont, Pascal ULiège Université de Liège - ULiège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Informatique et intelligence artificielle
    ORBi View his publications on ORBi
  • Total number of views 93
  • Total number of downloads 543










All documents available on MatheO are protected by copyright and subject to the usual rules for fair use.
The University of Liège does not guarantee the scientific quality of these students' works or the accuracy of all the information they contain.