Deep Learning and Ensemble Approach for Praise or Complaint Classification

Khedkar, Sujata; Shinde, Subhash K.

doi:10.1016/j.procs.2020.03.254

Cited by 27 publications

(17 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…To classify praise or complaint using linguistic-based hybrid features of extreme opinions, the authors in [15] that can automatically perform hotel reviews classification was introduced in [16] and provided an accuracy of 89% with 92% fi-score.…”

Section: Related Workmentioning

confidence: 99%

Towards Mining Public Opinion: An Attention-Based Long Short Term Memory Network Using Transfer Learning

Hossain,

Rashid,

Islam

et al. 2022

JCC

View full text Add to dashboard Cite

The Internet provides a large number of tools and resources, such as social media sites, online newsgroups, blogs, electronic forums, virtual communities, and online travel sites, for consumers to express their views or opinions regarding various issues. These opinions can help organizations like tourism to improve their products and services for their consumers. Opinion mining refers to a process of identifying emotions by applying Natural Language Processing (NLP) techniques to a pool of texts. This paper mainly focuses on mining public opinion from the hotel reviews domain. To do so, we proposed a novel technique called the Attention-Based Long Short Term Memory (Attention-LSTM) Network using a transfer learning approach. We empirically analyzed several machine learning and deep learning methods and observed our proposed technique provided an adequate performance for mining public opinion in the hotel reviews domain.

show abstract

Section: Related Workmentioning

confidence: 99%

Towards Mining Public Opinion: An Attention-Based Long Short Term Memory Network Using Transfer Learning

Hossain,

Rashid,

Islam

et al. 2022

JCC

View full text Add to dashboard Cite

show abstract

“…The information content of an alphabet is difficult to infer. Sujata and Shinde 22 highlighted that when selecting features its crucial to consider information content of the features. Therefore, the use of a higher level linguistic gram like a word or phrase is better since a word or a phrase contains more information than an alphabet.…”

Section: Related Workmentioning

confidence: 99%

“…Further the matrix generated was a 3‐dimensional matrix and the meaning of every word was considered thus leading to high computation cost and high features. Sujata and Shinde 22 used important words such as nouns, adjectives, intensifiers, verbs, and linguistic features of praise and complaint sentences and Affin dictionary to calculate extreme sentiment of each sentence. They further used Hybrid features like meta, Synthetic, content, and Semantic features of the sentences.…”

Section: Related Workmentioning

confidence: 99%

“…Although our work is in the area of feature extraction and selection for sentence level sentiment classification, it is slightly similar to the work in 14,15,19,22 since it uses word vector sentence representations using N‐gram and Lexicon dictionary. The key difference with our work is that while they use N‐grams, Word2Vec models and TF‐IDF to build document representations they do so from the entire document (sentence), in our work, we use tri‐ grams words, their POS tags and Semantic orientations identified from a specific part of a sentence by utilizing the proposals by references 25,26 .…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Lexicon‐pointed hybrid N‐gram Features Extraction Model (LeNFEM) for sentence level sentiment analysis

Mutinda

Mwangi

Okeyo

2021

Engineering Reports

View full text Add to dashboard Cite

Sentiment analysis of social media textual posts can provide information and knowledge that is applicable in social settings, business intelligence, evaluation of citizens' opinions in governance, and in mood triggered devices in the Internet of Things. Feature extraction and selection is a key determinant of accuracy and computational cost of machine learning models for such analysis. Most feature extraction and selection techniques utilize bag of words, N‐grams, and frequency‐based algorithms especially Term Frequency‐Inverse Document Frequency. However, these approaches do not consider relationships between words, they ignore words' characteristics and they suffer high feature dimensionality. In this paper we propose and evaluate a feature extraction and selection approach that utilizes a fixed hybrid N‐gram window for feature extraction and minimum redundancy maximum relevance feature selection algorithm for sentence level sentiment analysis. The approach improves the existing features extraction techniques, specifically the N‐gram by generating a hybrid vector from words, Part of Speech (POS) tags, and word semantic orientation. The vector is extracted by using a static trigram window identified by a lexicon where a sentiment word appears in a sentence. A blend of the words, POS tags, and the sentiment orientations of the static trigram are used to build the feature vector. The optimal features from the vector are then selected using minimum redundancy maximum relevance (MRMR) algorithm. Experiments were carried out using the public Yelp dataset to compare the performance of the proposed model and existing feature extraction models (BOW, normal N‐grams and lexicon‐based bag of words semantic orientations). Using supervised machine learning classifiers the experimental results showed that the proposed model had the highest F‐measure (88.64%) compared to the highest (83.55%) from baseline approaches. Wilcoxon test carried out ascertained that the proposed approach performed significantly better than the baseline approaches. Comparative performance analysis with other datasets further affirmed that the proposed approach is generalizable.

show abstract

“…Although our work is in the area of feature extraction and selection for sentence level sentiment classification, it is slightly similar to the work in 14,15,19,22 since it uses word vector sentence representations using N-gram and Lexicon dictionary. The key difference with our work is that while they use N-grams, Word2Vec models and TF-IDF to build document representations they do so from the entire document (sentence), in our work, we use tri-grams words, their POS tags and Semantic orientations identified from a specific part of a sentence by utilizing the proposals by references.…”

Section: Related Workmentioning

confidence: 99%

Lexicon - pointed hybrid N-gram Features Extraction Model (LeNFEM) for Sentence Level Sentiment Analysis.

Mutinda¹,

Mwangi²,

Okeyo³

2020

Preprint

View full text Add to dashboard Cite

Sentiment analysis of social media textual posts can provide information and knowledge that is applicable in social settings, business intelligence, evaluation of citizens' opinions in governance, and in mood triggered devices in the Internet of Things. Feature extraction and selection is a key determinant of accuracy and computational cost of machine learning models for such analysis. Most feature extraction and selection techniques utilize bag of words, N-grams, and frequency-based algorithms especially Term Frequency-Inverse Document Frequency. However, these approaches do not consider relationships between words, they ignore words' characteristics and they suffer high feature dimensionality. In this paper we propose and evaluate a feature extraction and selection approach that utilizes a fixed hybrid N-gram window for feature extraction and minimum redundancy maximum relevance feature selection algorithm for sentence level sentiment analysis. The approach improves the existing features extraction techniques, specifically the N-gram by generating a hybrid vector from words, Part of Speech (POS) tags, and word semantic orientation. The vector is extracted by using a static trigram window identified by a lexicon where a sentiment word appears in a sentence. A blend of the words, POS tags, and the sentiment orientations of the static trigram are used to build the feature vector. The optimal features from the vector are then selected using minimum redundancy maximum relevance (MRMR) algorithm. Experiments were carried out using the public Yelp dataset to compare the performance of the proposed model and existing feature extraction models (BOW, normal N-grams and lexicon-based bag of words semantic orientations). Using supervised machine learning classifiers the experimental results showed that the proposed model had the highest F-measure (88.64%) compared to the highest (83.55%) from baseline approaches. Wilcoxon test carried out ascertained that the proposed approach performed significantly better than the baseline approaches. Comparative performance analysis with other datasets further affirmed that the proposed approach is generalizable.

show abstract

Deep Learning and Ensemble Approach for Praise or Complaint Classification

Cited by 27 publications

References 23 publications

Towards Mining Public Opinion: An Attention-Based Long Short Term Memory Network Using Transfer Learning

Towards Mining Public Opinion: An Attention-Based Long Short Term Memory Network Using Transfer Learning

Lexicon‐pointed hybrid N‐gram Features Extraction Model (LeNFEM) for sentence level sentiment analysis

Lexicon - pointed hybrid N-gram Features Extraction Model (LeNFEM) for Sentence Level Sentiment Analysis.

Contact Info

Product

Resources

About