Adaptive two-stage feature selection for sentiment classification

Xu, Changxin; Siew, Tan Puay; Wang, Zhaoxia

doi:10.1109/smc.2017.8122782

Cited by 15 publications

(7 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…3.4.1 Naive Bayes Classifier (NB) It is based on Bayes' theorem ( 17), (35), which is a branch of probability theory. In this study, we employed one of the NB variants, Multinominal Naive Bayes (MNB), which treats the document as a probability of word order from the same word lexicon regardless of document length.…”

Section: Classificationmentioning

confidence: 99%

Influence of Diverse Feature Selection Methods in Sentiment Analysis

Madhusudhanan¹,

N.M²,

SenthilMurugan³

et al. 2022

ECS Trans.

View full text Add to dashboard Cite

With the enormous comments on social media and ecommerce, government, companies, and individuals eagerly want to know about people opinion about their policies, services, or any other. Sentiment analysis plays a major role to understand people emotions. Sentiment analysis alias opinion mining uses machine learning techniques to classify the given review as positive, negative, or neutral. This paper provides implementation of diverse feature selection techniques with different classifiers in sentiment analysis. The method is evaluated on various domains. Bag of Model with SVM provides satisfactory result on the datasets.

show abstract

Section: Classificationmentioning

confidence: 99%

Influence of Diverse Feature Selection Methods in Sentiment Analysis

Madhusudhanan¹,

N.M²,

SenthilMurugan³

et al. 2022

ECS Trans.

View full text Add to dashboard Cite

show abstract

“…Over the last two decades, few systems have been proposed to perform feature-based summarization. The summarizers are applied on various domains: product reviews [7], [14]- [17], movie reviews [4], local services reviews [18] and hotel reviews [2], [19], etc.…”

Section: A Nlp Techniques For Decision Making In E-commercementioning

confidence: 99%

“…We fine-tune BERT Base model to predict the sentiment orientation of the collected reviews. We build the model by creating a single new layer that will be trained with Large Movie Review Dataset v1.0 [64] 17 which contains 25,000 positive and 25,000 negative processed movie reviews. We set the sequence lenght to 128, the batch size to 32, the learning rate to 0.00002 and the number of epochs to 3.…”

Section: B Sentiment Analysismentioning

confidence: 99%

See 1 more Smart Citation

Aggregating Customer Review Attributes for Online Reputation Generation

Benlahbib

Nfaoui

2020

IEEE Access

View full text Add to dashboard Cite

In this paper, we face the problem of generating reputation for movies, products, hotels, restaurants and services by mining customer reviews expressed in natural language. To the best of our knowledge, previous studies on reputation generation for online entities have primarily examined semantic and sentiment orientation of customer reviews, disregarding other useful information that could be extracted from reviews, such as review helpfulness and review time. Therefore, we propose a new approach that combines review helpfulness, review time, review attached rating and review sentiment orientation for the purpose of generating a single reputation value toward various entities. The contribution of the paper is threefold. First, we design two equations to compute review helpfulness and review time scores, and we fine-tune Bidirectional Encoder Representations from Transformers (BERT) model to predict the review sentiment orientation probability. Second, we design a formula to assign a numerical score to each review. Then, we propose a new formula to compute reputation value toward the target entity (movie, product, hotel, restaurant, service, etc). Finally, we propose a new form to visualize reputation that depicts numerical reputation value, opinion categories, top positive review and top negative review. Experimental results coming from several real-world data sets of miscellaneous domains collected from IMDb, TripAdvisor and Amazon websites show the effectiveness of the proposed method in generating and visualizing reputation compared to three state-of-the-art reputation systems.

show abstract

“…Then, as a major contribution, the authors adapted methods from the econometrics literature, specifically the hedonic regression concept. As mentioned by Chi et al [56], existing feature selection techniques compute feature scores solely based on training data statistics or by modifying a specific feature metric formula to include test data information that cannot be generalized to other types of feature metrics, and they proposed combining both techniques (i.e., the training dataset and the feature metric formula). Mars and Gouider [57] proposed a big data architecture for decision-making, analyzing data and extracting customer opinions about product features.…”

Section: Introductionmentioning

confidence: 99%

Managing Marketing Decision-Making with Sentiment Analysis: An Evaluation of the Main Product Features Using Text Data Mining

et al. 2019

View full text Add to dashboard Cite

Companies have realized the importance of "big data" in creating a sustainable competitive advantage, and user-generated content (UGC) represents one of big data's most important sources. From blogs to social media and online reviews, consumers generate a huge amount of brand-related information that has a decisive potential business value for marketing purposes. Particularly, we focus on online reviews that could have an influence on brand image and positioning. Within this context, and using the usual quantitative star score ratings, a recent stream of research has employed sentiment analysis (SA) tools to examine the textual content of reviews and categorize buyer opinions. Although many SA tools split comments into negative or positive, a review can contain phrases with different polarities because the user can have different sentiments about each feature of the product. Finding the polarity of each feature can be interesting for product managers and brand management. In this paper, we present a general framework that uses natural language processing (NLP) techniques, including sentiment analysis, text data mining, and clustering techniques, to obtain new scores based on consumer sentiments for different product features. The main contribution of our proposal is the combination of price and the aforementioned scores to define a new global score for the product, which allows us to obtain a ranking according to product features. Furthermore, the products can be classified according to their positive, neutral, or negative features (visualized on dashboards), helping consumers with their sustainable purchasing behavior. We proved the validity of our approach in a case study using big data extracted from Amazon online reviews (specifically cell phones), obtaining satisfactory and promising results. After the experimentation, we could conclude that our work is able to improve recommender systems by using positive, neutral, and negative customer opinions and by classifying customers based on their comments.

show abstract

Adaptive two-stage feature selection for sentiment classification

Cited by 15 publications

References 24 publications

Influence of Diverse Feature Selection Methods in Sentiment Analysis

Influence of Diverse Feature Selection Methods in Sentiment Analysis

Aggregating Customer Review Attributes for Online Reputation Generation

Managing Marketing Decision-Making with Sentiment Analysis: An Evaluation of the Main Product Features Using Text Data Mining

Contact Info

Product

Resources

About