Titre :	How to analyze human sentiments : Covid19 pandemic
Type de document :	texte imprimé
Auteurs :	Ahmed Hechaichi, Auteur ; Mohamed Nadjib Mezerzi, Auteur ; Toumi,Lyazid, Directeur de thèse
Année de publication :	2022
Format :	29cm
Langues :	Français (fre)
Catégories :	Thèses & Mémoires:Informatique
Mots-clés :	Sentiment analysis Natural language processing
Index. décimale :	004 Informatique
Résumé :	Sentiment analysis, or opinion mining, is the computer examination of people’s sentiments, feelings, opinions, attitudes, moods, and emotions. It is one of the most active research fields in natural language processing, data mining, information retrieval, and web mining. Emotions and feelings have become more important in our daily lives in recent years. Many people assume that sentiment analysis is just the process of determining whether a document or a sentence reflects a positive or negative mood or opinion. It is, however, a far more complicated issue. In both business and society, this interesting subject is becoming more important. In this master’s thesis, we define the sentiment analysis topic and related concepts across the chapters, as well as the techniques used in preparing the collected dataset using nlp preprocessing techniques. We suggested and analyzed multiple machine learning and deep learning algorithms in order to categorize the tweets :naive bayes, support vector machine, logistic regression,k-nearest neighbor, random forest,decision tree, long short term memory, convolutional neural network ,hybrid of LSTM and CNN . The best results of the accuracy we have achieved was 90.21% with LSTM model in Corona_NLP dataset.
Côte titre :	MAI/0611
En ligne :	https://drive.google.com/file/d/1QizJGrkJ6tYVOQm0lIwOjE_bnsMlzTxd/view?usp=share [...]
Format de la ressource électronique :	pdf

How to analyze human sentiments : Covid19 pandemic [texte imprimé] / Ahmed Hechaichi, Auteur ; Mohamed Nadjib Mezerzi, Auteur ; Toumi,Lyazid, Directeur de thèse . - 2022 . -  ; 29cm.
Langues : Français (fre)

Catégories :	Thèses & Mémoires:Informatique
Mots-clés :	Sentiment analysis Natural language processing
Index. décimale :	004 Informatique
Résumé :	Sentiment analysis, or opinion mining, is the computer examination of people’s sentiments, feelings, opinions, attitudes, moods, and emotions. It is one of the most active research fields in natural language processing, data mining, information retrieval, and web mining. Emotions and feelings have become more important in our daily lives in recent years. Many people assume that sentiment analysis is just the process of determining whether a document or a sentence reflects a positive or negative mood or opinion. It is, however, a far more complicated issue. In both business and society, this interesting subject is becoming more important. In this master’s thesis, we define the sentiment analysis topic and related concepts across the chapters, as well as the techniques used in preparing the collected dataset using nlp preprocessing techniques. We suggested and analyzed multiple machine learning and deep learning algorithms in order to categorize the tweets :naive bayes, support vector machine, logistic regression,k-nearest neighbor, random forest,decision tree, long short term memory, convolutional neural network ,hybrid of LSTM and CNN . The best results of the accuracy we have achieved was 90.21% with LSTM model in Corona_NLP dataset.
Côte titre :	MAI/0611
En ligne :	https://drive.google.com/file/d/1QizJGrkJ6tYVOQm0lIwOjE_bnsMlzTxd/view?usp=share [...]
Format de la ressource électronique :	pdf