Diminishing or increasing? Exploring the impact of online movie reviews on consumer demand within and across two sequential release stages

This paper proposes an aspect based approach for sentiment analysis in Malayalam movie and product reviews. Sentiment Analysis is a cognitive process in which user's feeling and emotions are extracted. Now a day's people use social networking sites to discuss about movie reviews and product reviews so that internet itself can act as a recommendation system for its users. Thus social media monitoring, and VoC (Voice of the Customer) to track customer reviews, survey responses, competitors etc are becoming important day by day. Web has also become a medium for business analytics and situations in which text needs to be analyzed. Here comes the significance of Sentiment Analysis. Only polarity classification done by normal sentiment analysis task doesn't contribute much to the users. If the system also finds the aspect on which the user is commenting sentiment analysis task become more useful. The significance of aspect based sentiment analysis comes here. This system does sentence level aspect based sentiment analysis in Malayalam with 84.7% accuracy for Malayalam movie and product reviews.

Sentiment lexicons are useful to automatically extract sentiment from text. In this paper, we generate several Norwegian sentiment lexicons by extracting sentiment information from two different types of Norwegian text corpus, namely, news corpus and discussions forums. The methodology is based on the Point wise Mutual Information (PMI). We introduce a modification of the PMI that considers small blocks of the text instead of the text as a whole. The rational of this modification is to counter the detrimental effect of the length of the text on the PMI and on the quality of the lexicon. The high computational cost due to the huge amount of textual information to be processed in addition to our modified PMI formula is tackled efficiently by relying heavily on parallelization using Map-Reduce and Mongo DB shards. Movie and product reviews are used to evaluate the generated sentiment lexicons to correctly classify review ratings. The lexicon exhibits a satisfactory performance when evaluated, in particular when considering the context change in the corpuses. In fact, surprisingly enough, sentiment lexicon generated from large News corpus exhibits a satisfactory performance when tested on annotated product and movie reviews despite that the later two have different contexts, bearing similarity to the notion of Transfer Learning [1] reported in the literature. Some suggestions on how to increase the performance are proposed. All the sentiment lexicons are publicly available for those that are interested.

Many online retailers and some manufacturers/service providers have recently been engaging in questionable practices, where product reviews are often fabricated and/or posted without sufficient clarity and objectivity. Across an exploratory study and two main studies, we empirically examine this phenomenon and observe a pattern of effects that suggests that review valence (i.e., the average number of rating-stars a product receives) influences product attitudes and intentions, but that these outcomes are significantly impacted by the extent to which consumers are aware of potentially deceptive online review practices. Awareness of deceptive practices was found to differentially influence attitudes and intentions, depending upon whether the star-ratings were perfect (5/5 stars), highly positive (4.9/5 stars), or generally positive (4.5/5 or 4.7/5 stars). Participants’ perceptions of the e-retailer’s manipulative intent were also shown to mediate these effects, with higher perceptions of perceived manipulative intent yielding less favorable product attitudes and reduced purchase intentions.

Research problem: Increasingly, professional and technical communicators analyze, synthesize, and respond to user-generated content, including online consumer reviews of products, as the influence of user-generated content on consumers' purchasing decisions grows. But product reviews vary in the degree to which people perceive them to be credible. Research questions: (1) To what extent does a product review's environment—a retailer or brand site—affect review users' ratings of that review's credibility? (2) To what extent does review valence (positive versus negative) affect review users' ratings of review credibility? (3) What is the strength of the relationship among credibility and its two main components, trustworthiness and expertise? Literature review: Recent research has made clear the spread and the influence of user-generated comments and, thus, the need for sophistication in handling it. Review credibility has two main components: trustworthiness (which equates to honesty or sincerity) and expertise (which equates to accuracy). Prior research also shows the effects of valence (positivity or negativity) in reviews, noting that negative reviews have more influence than positive reviews on readers' perceptions of review credibility and purchasing decisions. Methodology: We tested the effect of a consumer review's environment (brand or retailer site) and the effect of review valence (positive or negative) on the perceived credibility of that review, as well the degree of correlation among credibility, trustworthiness, and expertise. Through an online survey, we exposed respondents to the same review text with different star ratings (4-star and 2-star) in two types of sites: brand and retailer. We asked participants to evaluate the review's credibility, trustworthiness, and expertise. In half of the exposures, participants evaluated a review in the site of a high-credibility company (Apple or Amazon), and in the other half of exposures, participants evaluated a review in the site of a midlevel-credibility company (Dell or Walmart). Results and conclusions: Credibility strongly correlated with both trustworthiness and expertise. Participants rated 4-star reviews as more credible than 2-star reviews on high-credibility sites, but star ratings had no impact on midlevel credibility sites. We found no difference between ratings of reviews displayed on brand and retailer sites for midlevel-credibility companies but a small difference between reviews displayed on brand and retailer sites for high-credibility companies. Professional communicators should attend to reviews posted both to retailer and brand sites. Conclusions: Professional communicators charged with managing user-generated content need not spend resources on channeling it into retailer and other independent review site environments as opposed to brand site environments. Our findings indicate that professional communicators looking to identify credible reviews should attend to review valence, or the positivity or negativity of a review. When managing user-generated product reviews, they should try to make credible content more noticeable to review users.

PurposeThe basic assumption is that there is a symmetric relationship between review valence and rating, but what if review valence and rating were linked asymmetrically? There are few studies which have investigated the situations in which positive and negative online reviews exert different influences on ratings. This study considers brand strength as having an important moderating role because the average rating of existing reviews for a particular product is a heuristic cue for decision makers. Thus, the purpose of this paper is to argue that an asymmetric relationship between review content valence and numerical rating will depend on brand strength.Design/methodology/approachThe authors have conducted a sentiment analysis via text mining, using self-developed computer programs to retrieve a data set from the TripAdvisor website.FindingsThis study finds there is an asymmetric relationship between review valence (verbal) and numerical rating. The authors further find brand strength to have an important moderating role. For a stronger brand, negative review content will have a greater impact on numerical ratings than positive review content, while for a weaker brand, positive review content will have a greater impact on numerical ratings than negative review content.Practical implicationsMarketers could adopt sentiment analysis via text mining of online reviews as a valid measure or predictor of consumer satisfaction or numerical ratings. Strong brands should direct more attention to negative reviews, because in such reviews the negative impact transcends the positive. In contrast, weak brands should aim to exploit as many positive reviews as possible to minimize the impact of any negative reviews.Originality/valueThis study finds there is an asymmetric relationship between review valence (verbal) and numerical rating and considers brand strength to play an important moderating role. The authors have used real data from the TripAdvisor website, which allow people to express themselves in an unsolicited manner, and linked these with the results from the sentiment analysis.

Purpose To explore the psychological mechanism through which consumer reviews affect people’s purchasing decisions and behavior, this study aims to examine the impact of statistical evidence embedded in product reviews on consumers’ perceptions and purchasing intentions. Design/methodology/approach The effects review valence and review volume are tested using a 3 (valence: positive vs neutral vs negative) × 2 (volume: high vs low) quasi-experimental design and online questionnaires. Findings The study finds that review valence has a stronger impact on consumers’ perceptions than review volume does. Negative reviews induce higher risk perception and a less favorable attitude toward purchases compared to positive reviews. In addition, although both attitude toward purchase and subjective norm are good antecedents of purchase intention, the attitude statistically has a stronger impact than the subjective norm. Research limitations/implications This study contributes to extant literature from three perspectives. The authors have reexamined the findings of econometric models and advanced their implications by explaining the related psychological changes in people’s perceptions. Second, the authors have extended the application of the theory of reasoned action and found it to be a good fit in explaining consumers’ behavior related to consumer reviews. And finally, the authors have provided a clear guideline on the magnitude of the effects of review valence and volume on consumers’ perceptions. Originality/value This study provides a good complement to econometric studies from both theoretical and practical perspectives. It bridges the gap between exploratory studies and behavioral studies in the field of consumer reviews.

Increasingly, user-generated product reviews serve as a valuable source of information for customers making product choices online. The existing literature typically incorporates the impact of product reviews on sales based on numeric variables representing the valence and volume of reviews. In this paper, we posit that the information embedded in product reviews cannot be captured by a single scalar value. Rather, we argue that product reviews are multifaceted, and hence the textual content of product reviews is an important determinant of consumers' choices, over and above the valence and volume of reviews. To demonstrate this, we use text mining to incorporate review text in a consumer choice model by decomposing textual reviews into segments describing different product features. We estimate our model based on a unique data set from Amazon containing sales data and consumer review data for two different groups of products (digital cameras and camcorders) over a 15-month period. We alleviate the problems of data sparsity and of omitted variables by providing two experimental techniques: clustering rare textual opinions based on pointwise mutual information and using externally imposed review semantics. This paper demonstrates how textual data can be used to learn consumers' relative preferences for different product features and also how text can be used for predictive modeling of future changes in sales.This paper was accepted by Ramayya Krishnan, information systems.

Sentiment analysis has been widely explored in many text domains, including tweets, movie reviews, shop/restaurant reviews, product reviews, and peer reviews for scholarly papers. However, it is very costly to manually label the training data for sentiment analysis. We focus on the problem and presents an approach for leveraging contextual features from unlabeled movie and restaurant reviews with a neural-network-based learning model, Ladder network. The experimental results by using two benchmark datasets, IMDb and YelpNYC, show that our model outperforms the baseline models including LSTM and SVM. Especially we verified that our model is better performance gaining on limited training datasets with 1% data labeled. Our source codes are available online.11Our source code can be obtained from https://github.com/jepyh/sentiment_analysis_few_labeled

Purpose To be sustainable and competitive in the current business environment, it is useful to understand users’ sentiment towards products and services. This critical task can be achieved via natural language processing and machine learning classifiers. This paper aims to propose a novel probabilistic committee selection classifier (PCC) to analyse and classify the sentiment polarities of movie reviews. Design/methodology/approach An Indian movie review corpus is assembled for this study. Another publicly available movie review polarity corpus is also involved with regard to validating the results. The greedy stepwise search method is used to extract the features/words of the reviews. The performance of the proposed classifier is measured using different metrics, such as F-measure, false positive rate, receiver operating characteristic (ROC) curve and training time. Further, the proposed classifier is compared with other popular machine-learning classifiers, such as Bayesian, Naïve Bayes, Decision Tree (J48), Support Vector Machine and Random Forest. Findings The results of this study show that the proposed classifier is good at predicting the positive or negative polarity of movie reviews. Its performance accuracy and the value of the ROC curve of the PCC is found to be the most suitable of all other classifiers tested in this study. This classifier is also found to be efficient at identifying positive sentiments of reviews, where it gives low false positive rates for both the Indian Movie Review and Review Polarity corpora used in this study. The training time of the proposed classifier is found to be slightly higher than that of Bayesian, Naïve Bayes and J48. Research limitations/implications Only movie review sentiments written in English are considered. In addition, the proposed committee selection classifier is prepared only using the committee of probabilistic classifiers; however, other classifier committees can also be built, tested and compared with the present experiment scenario. Practical implications In this paper, a novel probabilistic approach is proposed and used for classifying movie reviews, and is found to be highly effective in comparison with other state-of-the-art classifiers. This classifier may be tested for different applications and may provide new insights for developers and researchers. Social implications The proposed PCC may be used to classify different product reviews, and hence may be beneficial to organizations to justify users’ reviews about specific products or services. By using authentic positive and negative sentiments of users, the credibility of the specific product, service or event may be enhanced. PCC may also be applied to other applications, such as spam detection, blog mining, news mining and various other data-mining applications. Originality/value The constructed PCC is novel and was tested on Indian movie review data.

Online reviews have attracted much attention from firms as they play a significant role in consumers' purchase decisions. While prior investigations have explored the impact of online reviews on firms' operational decisions, little is known about how review volume and valence affect different players' operational decisions in a channel structure. We develop game-theoretic models to examine the effect of review volume and valence on different players' pricing strategies under two-period centralized, decentralized, and coordination structures composed of an online retailer and a manufacturer. The results indicate that the retailer and manufacturer benefit from a high review valence but are not necessarily harmed by a low review valence. Particularly, when review volume is sufficiently large, the low review valence may also bring more profits. In this case, online reviews can expand the potential market. Finally, a two-period two-part tariff contract can perfectly coordinate the supply chain and create a win-win situation for both the retailer and manufacturer under certain conditions. The manufacturer may charge a sufficiently low first-period wholesale price and even subsidize the retailer under a low valence and moderate number of reviews. Our results offer a more complete understanding of the implications of reviews in supply chain management.

Sentiment Classification is a fundamental task in the field of Natural Language Processing, and has very important academic and commercial applications. It aims to automatically predict the degree of sentiment present in a text that contains opinions and subjectivity at some level, like product and movie reviews, or tweets. This can be really difficult to accomplish, in part, because different domains of text contains different words and expressions. In addition, this difficulty increases when text is written in a non-English language due to the lack of databases and resources. As a consequence, several cross-domain and cross-language techniques are often applied to this task in order to improve the results. In this work we perform a study on the ability of a classification system trained with a large database of product reviews to generalize to different Spanish domains. Reviews were collected from the MercadoLibre website from seven Latin American countries, allowing the creation of a large and balanced dataset. Results suggest that generalization across domains is feasible though very challenging when trained with these product reviews, and can be improved by pre-training and fine-tuning the classification model.

Sentiment classification is to find the polarity of product or user reviews. Supervised machine learning algorithms is used for opinion mining such as naive Bayes, K-nearest neighbour, decision trees, maximum entropy and hidden Markov model and support vector machine. KNN is a simple algorithm, but a less efficient classification algorithm. In this paper, we propose an improved KNN algorithm. An optimised feature selection, genetic algorithm that incorporates the information gain for feature selection and combined with bagging technique and KNN for improving the accuracy of sentiment classification. Specifically, we compared two approaches and traditional KNN for sentiment classification of movie reviews and product reviews. The same approach has been applied to other machine learning algorithms such as support vector machine and naive Bayes and the result is compared with POS-based feature set method. The proposed method is evaluated and experimental results using information gain, genetic algorithm with bagging technique indicate higher performance result with accuracy of 87.50% of the movie reviews and exhibits better performance in terms of accuracy, precision and recall for movie, DVD, electronics and kitchen reviews.

소수 의견을 포함하는 온라인 상품평은 긍정 또는 부정 일변도인 상품평에서는 찾기 어려운 유익한 정보를 내포하기도 한다. 본 논문에서는 주어진 상품평 집합 속에서 소수상품평을 검색하는 방법을 제안한다. 제안방법은 개별 상품평을 먼저 긍정/부정 상품평으로 자동분류한 뒤, 주어진 상품평 집합의 긍정/부정 상품평의 비대칭도를 계산하여 소수상품평을 검색한다. 소수상품평 검색에서는 긍정/부정 자동분류 성능이 소수상품평 검색성능에 영향을 주는데, 본 논문에서는 도메인에 특화된 감성사전과 그렇지 않은 일반적인 감성사전을 가지고 상품평을 긍정/부정으로 감성분류한 뒤 비대칭도를 계산하여 소수상품평 검색성능을 비교한다. 스마트폰과 영화를 다룬 온라인 영문 상품평에 대하여 도메인에 특화된 감성사전을 가지고 소수상품평 검색성능을 평가한 결과, F1점수는 각각 24.6%와 15.9%였고, 정확도는 각각 56.8%와 46.8%였다. 이는 스마트폰과 영화의 개별 상품평 긍정/부정 분류 정확도가 각각 85.3%와 78.8%일 때의 성능이다. 본 논문에서는 또 긍정/부정 자동분류 성능이 주어졌을 때의 이론적인 소수상품평 검색성능에 대해서도 논의한다. A given product's online product reviews build up to form largely positive or negative reviews or mixed reviews that include both the positive and negative reviews. While the homogeneously positive or negative reviews help readers identify the generally praised or criticized product, the mixed reviews with minority opinions potentially contain valuable information about the product. We present a method of retrieving minority opinions from the online product reviews using the skewness of positive/negative reviews. The proposed method first classifies the positive/negative product reviews using a sentiment dictionary and then calculates the skewness of the classified results to identify minority reviews. Minority review retrieval experiments were conducted on smartphone and movie reviews, and the F1-measures were 24.6% (smartphone) and 15.9% (movie) and the accuracies were 56.8% and 46.8% when the individual reviews' sentiment classification accuracies were 85.3% and 78.8%. The theoretical performance of minority review retrieval is also discussed.

This research investigates the joint impact of numeric ratings and review sentiments in prior product reviews on consumers’ product evaluations and how this impact varies between individualist and collectivist cultures. Using data on 115,231 consumer reviews of 167 American movies released both in the US and China, the authors find that the joint impact of movie ratings and review sentiments varies with the consistency between the two information sources, as well as their valence. Furthermore, their analysis reveals that the joint impact differs between the two cultures. The authors suggest that these findings may arise because consumers’ perception of information credibility varies depending on the information consistency and their cultural background. The findings offer important implications for firms in managing UGC platforms for consumers’ product reviews and social media across global markets.

How product review voting is influenced by existing votes, consumer involvement, review valence, and review diagnosticity