Customized reputation generation of entities using sentiment analysis

2020 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Arpita Gupta ◽  
Saloni Priyani ◽  
Ramadoss Balakrishnan
Keyword(s):  
Sentiment Analysis ◽  
Classification Models ◽  
Data Set ◽  
Content Type ◽  
Customer Reviews ◽  
Art Methods ◽  
Review Helpfulness ◽  
Sentiment Score ◽  
State Of Art

Purpose In this study, the authors have used the customer reviews of books and movies in natural language for the purpose of sentiment analysis and reputation generation on the reviews. Most of the existing work has performed sentiment analysis and reputation generation on the reviews by using single classification models and considered other attributes for reputation generation. Design/methodology/approach The authors have taken review, helpfulness and rating into consideration. In this paper, the authors have performed sentiment analysis for extracting the probability of the review belonging to a class, which is further used for generating the sentiment score and reputation of the review. The authors have used pre-trained BERT fine-tuned for sentiment analysis on movie and book reviews separately. Findings In this study, the authors have also combined the three models (BERT, Naïve Bayes and SVM) for more accurate sentiment classification and reputation generation, which has outperformed the best BERT model in this study. They have achieved the best accuracy of 91.2% for the movie review data set and 89.4% for the book review data set which is better than the existing state-of-art methods. They have used the transfer learning concept in deep learning where you take knowledge gained from one problem and apply it to a similar problem. Originality/value The authors have proposed a novel model based on combination of three classification models, which has outperformed the existing state-of-art methods. To the best of the authors’ knowledge, there is no existing model which combines three models for sentiment score calculation and reputation generation for the book review data set.

Reviews’ length and sentiment as correlates of online reviews’ ratings

2018 ◽  
Vol 28 (3) ◽  
pp. 544-563 ◽  
Author(s):  
Maryam Ghasemaghaei ◽  
Seyed Pouyan Eslami ◽  
Ken Deal ◽  
Khaled Hassanein
Keyword(s):  
Sentiment Analysis ◽  
Online Reviews ◽  
Marketing Strategies ◽  
Insurance Companies ◽  
Online Review ◽  
Data Set ◽  
Content Type ◽  
Customer Reviews ◽  
Service Companies ◽  
Consumer Review

Purpose The purpose of this paper is twofold: first, to identify and validate reviews’ length and sentiment as correlates of online reviews’ ratings; and second, to understand the emotions embedded in online reviews and how they associate with specific words used in such reviews. Design/methodology/approach A panel data set of customer reviews was collected for auto, life, and home insurance from January 2012 to December 2015 using a web scraping technique. Using a sentiment analysis approach, 1,584 reviews for the auto, home, and life insurance services of 156 insurance companies were analyzed. Findings The results indicate that, since 2013, consumers have generally had more negative emotions than positive ones toward insurance services. The results also show that consumer review sentiment correlates positively and review length correlates negatively with consumer online review ratings. Furthermore, a two-way ANOVA analysis shows that, in general, short reviews with positive sentiment are associated with high review ratings. Practical implications The findings of this study provide service companies, in general, and insurance companies, in particular, with important guidelines that should be considered to increase consumers’ positive attitude toward their services. Originality/value This paper highlights the importance of sentiment analysis in identifying consumer reviews’ emotions and understanding the associations and interactions of reviews’ length and sentiment on online review rating, which can lead to improved marketing strategies.

Do online consumer reviews help to evaluate the performance of automobile manufacturers?

2020 ◽  
Vol 33 (5) ◽  
pp. 1153-1198
Author(s):  
Amit Singh ◽  
Mamata Jenamani ◽  
Jitesh Thakkar
Keyword(s):  
Sentiment Analysis ◽  
Data Set ◽  
Pareto Chart ◽  
Content Type ◽  
Customer Reviews ◽  
Online Consumer Reviews ◽  
Online Customer Reviews ◽  
Fishbone Diagram ◽  
Online Consumer ◽  
Practical Implications

PurposeThis research proposes a text analytics–based framework that examines the utility of online customer reviews in evaluating automobile manufacturers and discovering their consumer-perceived weaknesses.Design/methodology/approachThe proposed framework integrates aspect-level sentiment analysis with the house of quality (HoQ), TOPSIS, Pareto chart and fishbone diagram. While sentiment analysis mines and quantifies review-embedded consumer opinions on various automobile attributes, the integrated HoQ-TOPSIS analyzes the quantified opinions and evaluates the manufacturers. The Pareto charts assist in discovering consumer-perceived weaknesses of the underperforming manufacturers. Finally, the fishbone diagram visually represents the results in the form with which the manufacturing community is acquainted.FindingsThe proposed framework is tested on a review data set collected from CarWale, a well-known car portal in India. Selecting five manufacturers from the mid-size car segment, the authors identified the worst-performing one and discovered its weak attributes.Practical implicationsThe proposed framework can help the manufacturers in evaluating competitor; identifying consumers' contemporary interests; discovering own and their competitors' weak attributes; assessing the suppliers and sending early warnings; detecting the hazardous defects. It can assist the component suppliers in devising process improvement strategies; improving their customer network; comparing them with competitors. It can support the customers in identifying the best available alternative.Originality/valueThe proposed framework is first of its kind to integrate the sentiment analysis with (1) HoQ-TOPSIS to assess the manufacturers; (2) Pareto chart to discover their weaknesses; (3) fishbone diagram to visually represent the results.

Determining banking service attributes from online reviews: text mining and sentiment analysis

2022 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Divya Mittal ◽  
Shiv Ratan Agrawal
Keyword(s):  
Text Mining ◽  
Customer Satisfaction ◽  
Sentiment Analysis ◽  
Banking Sector ◽  
Online Reviews ◽  
Socioeconomic Background ◽  
Content Type ◽  
Customer Reviews ◽  
Banking Service ◽  
The Interest Rate

PurposeThe current study employs text mining and sentiment analysis to identify core banking service attributes and customer sentiment in online user-generated reviews. Additionally, the study explains customer satisfaction based on the identified predictors.Design/methodology/approachA total of 32,217 customer reviews were collected across 29 top banks on bankbazaar.com posted from 2014 to 2021. In total three conceptual models were developed and evaluated employing regression analysis.FindingsThe study revealed that all variables were found to be statistically significant and affect customer satisfaction in their respective models except the interest rate.Research limitations/implicationsThe study is confined to the geographical representation of its subjects' i.e. Indian customers. A cross-cultural and socioeconomic background analysis of banking customers in different countries may help to better generalize the findings.Practical implicationsThe study makes essential theoretical and managerial contributions to the existing literature on services, particularly the banking sector.Originality/valueThis paper is unique in nature that focuses on banking customer satisfaction from online reviews and ratings using text mining and sentiment analysis.

Aspect-Based Sentiment Analysis of Online Product Reviews

2017 ◽  
pp. 175-191 ◽  
Author(s):  
Vinod Kumar Mishra ◽  
Himanshu Tiruwa
Keyword(s):  
Sentiment Analysis ◽  
Computational Linguistics ◽  
Language Processing ◽  
Future Trend ◽  
Product Reviews ◽  
Customer Reviews ◽  
Part Of Speech ◽  
Lexical Approach ◽  
Sentiment Score ◽  
Pos Tagger

Sentiment analysis is a part of computational linguistics concerned with extracting sentiment and emotion from text. It is also considered as a task of natural language processing and data mining. Sentiment analysis mainly concentrate on identifying whether a given text is subjective or objective and if it is subjective, then whether it is negative, positive or neutral. This chapter provide an overview of aspect based sentiment analysis with current and future trend of research on aspect based sentiment analysis. This chapter also provide a aspect based sentiment analysis of online customer reviews of Nokia 6600. To perform aspect based classification we are using lexical approach on eclipse platform which classify the review as a positive, negative or neutral on the basis of features of product. The Sentiwordnet is used as a lexical resource to calculate the overall sentiment score of each sentence, pos tagger is used for part of speech tagging, frequency based method is used for extraction of the aspects/features and used negation handling for improving the accuracy of the system.

scholarly journals Translating emotional insights from hospitality employees’ comments

2018 ◽  
Vol 32 (1) ◽  
pp. 75-92 ◽  
Author(s):  
Lisa M. Young ◽  
Swapnil Rajendra Gavade
Keyword(s):  
Job Satisfaction ◽  
Sentiment Analysis ◽  
Big Data Analytics ◽  
Large Data ◽  
Hospitality Management ◽  
Data Set ◽  
Employee Survey ◽  
Content Type ◽  
Employee Job Satisfaction ◽  
The Usa

PurposeThe purpose of this paper is to use the data analysis method of sentiment analysis to improve the understanding of a large data set of employee comments from an annual employee job satisfaction survey of a US hospitality organization.Design/methodology/approachSentiment analysis is used to examine the employee comments by identifying meaningful patterns, frequently used words and emotions. The statistical computing language, R, uses the sentiment analysis process to scan each employee survey comment, compare the words with the predefined word dictionary and classify the employee comments into the appropriate emotion category.FindingsEmployee responses written in English and in Spanish are compared with significant differences identified between the two groups, triggering further investigation of the Spanish comments. Sentiment analysis was then conducted on the Spanish comments comparing two groups, front-of-house vs back-of-house employees and employees with male supervisors vs female supervisors. Results from the analysis of employee comments written in Spanish point to higher scores for job sadness and anger. The negative comments referred to desires for improved healthcare, requests for increased wages and frustration with difficult supervisor relationships. The findings from this study add to the growing body of literature that has begun to focus on the unique work experiences of Latino employees in the USA.Originality/valueThis is the first study to examine a large unstructured English and Spanish text database from a hospitality organization’s employee job satisfaction surveys using sentiment analysis. Applying this big data analytics process to advance new insights into the human capital aspects of hospitality management is intriguing to many researchers. The results of this study demonstrate an issue that needs to be further investigated particularly considering the hospitality industry’s employee demographics.

Product weakness finder: an opinion-aware system through sentiment analysis

2014 ◽  
Vol 114 (8) ◽  
pp. 1301-1320 ◽  
Author(s):  
Hongwei Wang ◽  
Wei Wang
Keyword(s):  
Sentiment Analysis ◽  
Comparative Evaluation ◽  
Analytical Framework ◽  
Online Reviews ◽  
Product Reviews ◽  
Network Algorithms ◽  
Content Type ◽  
Sentiment Mining ◽  
Highly Correlated ◽  
Sentiment Score

Purpose – Extant methods of product weakness detection usually depend on time-consuming questionnaire with high artificial involvement, so the efficiency and accuracy are not satisfied. The purpose of this paper is to propose an opinion-aware analytical framework – PRODWeakFinder – to expect to detect product weaknesses through sentiment analysis in an effective way. Design/methodology/approach – PRODWeakFinder detects product weakness by considering both comparative and non-comparative evaluations in online reviews. For comparative evaluation, an aspect-oriented comparison network is built, and the authority is assessed for each node by network analysis. For non-comparative evaluation, sentiment score is calculated through sentiment analysis. The composite score of aspects is calculated by combing the two types of evaluations. Findings – The experiments show that the comparative authority score and the non-comparative sentiment score are not highly correlated. It also shows that PRODWeakFinder outperforms the baseline methods in terms of accuracy. Research limitations/implications – Semantic-based method such as ontology are expected to be applied to identify the implicit features. Furthermore, besides PageRank, other sophisticated network algorithms such as HITS will be further employed to improve the framework. Practical implications – The link-based network is more suitable for weakness detection than the weight-based network. PRODWeakFinder shows the potential on reducing overall costs of detecting product weaknesses for companies. Social implications – A quicker and more effective way would be possible for weakness detection, enabling to reduce product defects and improve product quality, and thus raising the overall social welfare. Originality/value – An opinion-aware analytical framework is proposed to sentiment mining of online product reviews, which offer important implications regarding how to detect product weaknesses.

Predicting employee attrition using tree-based models

2020 ◽  
Vol 28 (6) ◽  
pp. 1273-1291
Author(s):  
Nesreen El-Rayes ◽  
Ming Fang ◽  
Michael Smith ◽  
Stephen M. Taylor
Keyword(s):  
Random Forest ◽  
Decision Tree ◽  
Prediction Models ◽  
Binary Classification ◽  
Primary Role ◽  
Classification Models ◽  
Job Transition ◽  
Data Set ◽  
Content Type ◽  
Employee Attrition

Purpose The purpose of this study is to develop tree-based binary classification models to predict the likelihood of employee attrition based on firm cultural and management attributes. Design/methodology/approach A data set of resumes anonymously submitted through Glassdoor’s online portal is used in tandem with public company review information to fit decision tree, random forest and gradient boosted tree models to predict the probability of an employee leaving a firm during a job transition. Findings Random forest and decision tree methods are found to be the strongest attrition prediction models. In addition, compensation, company culture and senior management performance play a primary role in an employee’s decision to leave a firm. Practical implications This study may be used by human resources staff to better understand factors which influence employee attrition. In addition, techniques developed in this study may be applied to company-specific data sets to construct customized attrition models. Originality/value This study contains several novel contributions which include exploratory studies such as industry job transition percentages, distributional comparisons between factors strongly contributing to employee attrition between those who left or stayed with the firm and the first comprehensive search over binary classification models to identify which provides the strongest predictive performance of employee attrition.

A generalizable sentiment analysis method for creating a hotel dictionary: using big data on TripAdvisor hotel reviews

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Sayeh Bagherzadeh ◽  
Sajjad Shokouhyar ◽  
Hamed Jahani ◽  
Marianna Sigala
Keyword(s):  
Machine Learning ◽  
Big Data ◽  
Sentiment Analysis ◽  
Learning Algorithm ◽  
Tourism Industry ◽  
Bag Of Words ◽  
Machine Learning Algorithm ◽  
End User ◽  
Content Type ◽  
Customer Reviews

Purpose Research analyzing online travelers’ reviews has boomed over the past years, but it lacks efficient methodologies that can provide useful end-user value within time and budget. This study aims to contribute to the field by developing and testing a new methodology for sentiment analysis that surpasses the standard dictionary-based method by creating two hotel-specific word lexicons. Design/methodology/approach Big data of hotel customer reviews posted on the TripAdvisor platform were collected and appropriately prepared for conducting a binary sentiment analysis by developing a novel bag-of-words weighted approach. The latter provides a transparent and replicable procedure to prepare, create and assess lexicons for sentiment analysis. This approach resulted in two lexicons (a weighted lexicon, L1 and a manually selected lexicon, L2), which were tested and validated by applying classification accuracy metrics to the TripAdvisor big data. Two popular methodologies (a public dictionary-based method and a complex machine-learning algorithm) were used for comparing the accuracy metrics of the study’s approach for creating the two lexicons. Findings The results of the accuracy metrics confirmed that the study’s methodology significantly outperforms the dictionary-based method in comparison to the machine-learning algorithm method. The findings also provide evidence that the study’s methodology is generalizable for predicting users’ sentiment. Practical implications The study developed and validated a methodology for generating reliable lexicons that can be used for big data analysis aiming to understand and predict customers’ sentiment. The L2 hotel dictionary generated by the study provides a reliable method and a useful tool for analyzing guests’ feedback and enabling managers to understand, anticipate and re-actively respond to customers’ attitudes and changes. The study also proposed a simplified methodology for understanding the sentiment of each user, which, in turn, can be used for conducting comparisons aiming to detect and understand guests’ sentiment changes across time, as well as across users based on their profiles and experiences. Originality/value This study contributes to the field by proposing and testing a new methodology for conducting sentiment analysis that addresses previous methodological limitations, as well as the contextual specificities of the tourism industry. Based on the paper’s literature review, this is the first research study using a bag-of-words approach for conducting a sentiment analysis and creating a field-specific lexicon.

Lazy collaborative filtering with dynamic neighborhoods

2018 ◽  
Vol 46 (2) ◽  
pp. 95-109
Author(s):  
Suganeshwari G. ◽  
Syed Ibrahim S.P. ◽  
Gang Li
Keyword(s):  
Collaborative Filtering ◽  
Recommendation System ◽  
Temporal Information ◽  
High Quality ◽  
Good Efficiency ◽  
Content Type ◽  
Active User ◽  
Art Methods ◽  
Information Recommendation ◽  
State Of Art

Purpose The purpose of this paper is to address the scalability issue and produce high-quality recommendation that best matches the user’s current preference in the dynamically growing datasets in the context of memory-based collaborative filtering methods using temporal information. Design/methodology/approach The proposed method is formalized as time-dependent collaborative filtering method. For each item, a set of influential neighbors is identified by using the truncated version of similarity computation based on the timestamp. Then, recent n transactions are used to generate the recommendation that reflect the recent preference of the active user. The proposed method, lazy collaborative filtering with dynamic neighborhoods (LCFDN), is further scaled up by implementing in spark using parallel processing paradigm MapReduce. The experiments conducted on MovieLens dataset reveal that LCFDN implemented on MapReduce is more efficient and achieves good performance than the existing methods. Findings The results of the experimental study clearly show that not all ratings provide valuable information. Recommendation system based on LCFDN increases the efficiency of predictions by selecting the most influential neighbors based on the temporal information. The pruning of the recent transactions of the user also addresses the user’s preference drifts and is more scalable when compared to state-of-art methods. Research limitations/implications In the proposed method, LCFDN, the neighborhood space is dynamically adjusted based on the temporal information. In addition, the LCFDN also determines the user’s current interest based on the recent preference or purchase details. This method is designed to continuously track the user’s preference with the growing dataset which makes it suitable to be implemented in the e-commerce industry. Compared with the state-of-art methods, this method provides high-quality recommendation with good efficiency. Originality/value The LCFDN is an extension of collaborative filtering with temporal information used as context. The dynamic nature of data and user’s preference drifts are addressed in the proposed method by dynamically adapting the neighbors. To improve the scalability, the proposed method is implemented in big data environment using MapReduce. The proposed recommendation system provides greater prediction accuracy than the traditional recommender systems.

Recommending valuable ideas in an open innovation community

2018 ◽  
Vol 118 (4) ◽  
pp. 683-699 ◽  
Author(s):  
Hanjun Lee ◽  
Keunho Choi ◽  
Donghee Yoo ◽  
Yongmoo Suh ◽  
Soowon Lee ◽  
...  
Keyword(s):  
Open Innovation ◽  
Recommendation System ◽  
Information Overload ◽  
Idea Generation ◽  
Large Data ◽  
Classification Models ◽  
Data Set ◽  
Content Type ◽  
Strategic Assets ◽  
Innovation Community

PurposeOpen innovation communities are a growing trend across diverse industries because they provide opportunities of collaborating with customers and exploiting their knowledge effectively. Although open innovation communities can be strategic assets that can help firms innovate, firms nonetheless face the challenge of information overload incurred due to the characteristic of the community. The purpose of this paper is to mitigate the problem of information overload in an open innovation environment.Design/methodology/approachThis study chose MyStarbucksIdea.com (MSI) as a target open innovation community in which customers share their ideas. The authors analyzed a large data set collected from MSI utilizing text mining techniques including TF-IDF and sentiment analysis, while considering both term and non-term features of the data set. Those features were used to develop classification models to calculate the adoption probability of each idea.FindingsThe results showed that term and non-term features play important roles in predicting the adoptability of ideas and the best classification accuracy was achieved by the hybrid classification models. In most cases, the precisions of classification models decreased as the number of recommendations increased, while the models’ recalls and F1s increased.Originality/valueThis research dealt with the problem of information overload in an open innovation context. A large amount of customer opinions from an innovation community were examined and a recommendation system to mitigate the problem was proposed. Using the proposed system, the firm can get recommendations for ideas that could be valuable for its business innovation in the idea generation phase, thereby resolving the information overload and enhancing the effectiveness of open innovation.

Sign in / Sign up

Export Citation Format

Share Document