Multi-dimensional Sentiment Analysis for Large-Scale E-commerce Reviews

Zheng, Lizhou; Jin, Peiquan; Zhao, Jie; Yue, Lihua

doi:10.1007/978-3-319-10085-2_41

Cited by 8 publications

(3 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The results produced are better as compared to previous studies (Kim, Pantel et al (2006); Liu, Cao et al (2007)). Zheng, Jin et al (2014) presented a multi-dimensional sentiment analysis approach and propose an algorithm for sentiment analysis based on rules and a dimensional sentiment lexicon. The results demonstrated that proposed method outperforms the traditional approaches.…”

Section: Consumer Sentiments Emotion Features and Review Helpfulnessmentioning

confidence: 99%

Helpfulness of product reviews as a function of discrete positive and negative emotions

Malik

Hussain

2017

Computers in Human Behavior

141

View full text Add to dashboard Cite

The product review plays an important role in customer's purchase decision making process on the e-commerce websites. Emotions can significantly influence the way that reviews are processed. The importance of discrete emotions embedded in online reviews and their impact on review helpfulness is not explored intensively in prior studies. This study builds a helpfulness predictive model using deep neural network and investigates the influences of emotions that contribute to review helpfulness. We present an approach that extract novel discrete positive and negative emotion features from textual content of product reviews using NRC emotion Lexicon. In addition, the type of product, reviewer, visibility, readability, linguistics and sentiment related characteristics are also used for comparison and helpfulness prediction. The experimental results on two real-life datasets demonstrate that positive emotion features are the best predictors when individual category of feature is considered. However, negative emotion features and visibility features have comparable performance. Furthermore, the hybrid set of features with positive emotion features produce the best predictive performance for helpfulness of online reviews. The empirical evaluation finds that Trust, Joy and Anticipation (positive emotions); Anxiety and Sadness (negative emotions) are most influential emotion dimensions and have greater impact on perceived helpfulness. The findings of this study highlight the importance of emotions in online reviews and have significant implications for consumer and ecommerce retailers.

show abstract

Section: Consumer Sentiments Emotion Features and Review Helpfulnessmentioning

confidence: 99%

Helpfulness of product reviews as a function of discrete positive and negative emotions

Malik

Hussain

2017

Computers in Human Behavior

141

View full text Add to dashboard Cite

show abstract

“…Thus, people can expand their network in society and finally form a big social network. Generally, social networks can provide a number of new applications and services, such as news feature exploration [7], event monitoring [8,9], and sentiment analysis of reviews [10].…”

Section: Knowledge Sharing In Social Networkmentioning

confidence: 99%

User Willingness toward Knowledge Sharing in Social Networks

et al. 2018

Self Cite

View full text Add to dashboard Cite

Social networks introduce new potential for people to share knowledge with others. However, it is not clear what factors influence user willingness toward knowledge sharing in social networks. Aiming to answer these questions, in this paper we analyze the major factors influencing user willingness toward knowledge sharing in social networks and propose a new research model that is inspired by the technology acceptance model (TAM). In particular, we introduce a new independent variable called perceived value which is described by four aspects: social value, entertainment value, emotion value, and information value. In addition, we introduce a new mediating variable, trust, to reflect the intermediating relationship between perceived value and knowledge-sharing willingness. We conduct an empirical analysis on questionnaire data and present comprehensive results on reliability and validity, factor analysis, correlation analysis, and mediating effects analysis. The results show that perceived value has a significant impact on knowledge-sharing willingness, and trust plays a partial intermediate role between perceived value and knowledge-sharing willingness. Further, we present some research implications for knowledge sharing and learning innovation in social networks, as well as some suggestions for organizations to advance knowledge sharing and learning innovation in the social-network age.

show abstract

“…This predicament is compounded by the sheer profusion of comments that, often, inundates consumers, prompting them to restrict their perusal to the preliminary comments on a product's review page. This myopic approach, however, inadvertently relegates a substantial corpus of valuable insights to obscurity, beyond the purview of prospective purchasers [15], [16].…”

Section: Article Historymentioning

confidence: 99%

Advanced product review summarization in e-commerce marketplaces: elevating beyond tf-idf and lexrank method

Anggraeni,

Ismi

2022

SITech

View full text Add to dashboard Cite

In the fiercely competitive domain of online product sales, wherein engendering trust among prospective buyers assumes paramount significance, the role of product reviews cannot be understated. However, a prevailing issue in online marketplaces resides in the presence of product reviews that do not consistently align with the overall product rating. Furthermore, the sheer abundance of comments often leads potential consumers to confine their scrutiny to the initial comments, thus leaving a substantial volume of reviews unexplored. To rectify this challenge, this study introduces an automated text summarization system for product reviews, leveraging the LexRank methodology. This system underwent rigorous evaluation using the Rouge metric, with results manifesting substantial promise. At a threshold of 0.1, Rouge-1 exhibited an accuracy of 16.67%, while Rouge-2 scored 3.01%, and Rouge-L reached 16.50%. At a threshold of 0.2, Rouge-1 yielded a score of 16.08%, Rouge-2 registered 2.64%, and Rouge-L scored 16.57%. The second evaluation, performed with a distinct test dataset, notably excelled, emphasizing the system's competence. Specifically, at the 0.2 threshold, the system displayed superior performance, underscoring its efficacy in refining product review summarization within online marketplaces

show abstract

Multi-dimensional Sentiment Analysis for Large-Scale E-commerce Reviews

Cited by 8 publications

References 22 publications

Helpfulness of product reviews as a function of discrete positive and negative emotions

Helpfulness of product reviews as a function of discrete positive and negative emotions

User Willingness toward Knowledge Sharing in Social Networks

Advanced product review summarization in e-commerce marketplaces: elevating beyond tf-idf and lexrank method

Contact Info

Product

Resources

About