In the buying decision process, online reviews become an important source of information. They become the basis of evaluating alternatives before making purchase decision. This paper proposes a methodology to reveal one of the hidden alternative evaluation processes by identifying the relation between the observable online customer reviews and sales rank. This methodology applies a combined approach of word embedding (word2vec) and X-means clustering, which produces product-feature words. It is followed by identifying sentiment words and their intensity, determining connection of words from dependency tree, and finally relating variables from the reviews to the sales rank of a product by a regression model. The methodology is applied to two data sets of wearable technology and laptop products. As implied by the high predicted R-squared values, the models are generalizable into new data sets. Among the interesting findings are the statements of problems or issues of a product are related to better sales rank, and many product features that are mentioned in the review title are significantly related to sales rank. For product designers, the significant variables in the regression models suggest the possible product features to be improved.
This paper proposes a data-driven methodology to automatically identify product usage contexts from online customer reviews. Product usage context is one of the factors that affect product design, consumer behavior, and consumer satisfaction. The previous works identify the usage contexts using the survey-based method or subjectively determine them. The proposed methodology, on the other hand, uses machine learning and Natural Language Processing tools to identify and cluster usage contexts from a large volume of customer reviews. Furthermore, aspect sentiment analysis is applied to capture the sentiment toward a particular usage context in a sentence. The methodology is implemented to two data sets of products, i.e., laptop and tablet. The result shows that the methodology is able to capture relevant product usage contexts and cluster bigrams that refer to similar usage context. The aspect sentiment analysis enables the observation of a product’s position with respect to its competitors for a particular usage context. For a product designer, the observation may indicate a requirement to improve the product. It may also indicate a possible market opportunity in a usage context in which most of the current products are perceived negatively by customers. Finally, it is shown that overall rating might not be a strong indicator for representing customer sentiment toward a particular usage context, due to the moderate linear correlation for most of the usage contexts in the case study.
The recent development in engineering design has incorporated customer preferences by involving a choice model. In generating a choice model to produce a good quality estimate of parameters related to product attributes, a high-quality choice set is essential. However, the choice set data are often not available. This research proposes a methodology that utilizes online data and customer reviews to construct customer choice sets in the absence of both the actual choice set and the customer sociodemographic data. The methodology consists of three main parts, i.e., clustering the products based on their attributes, clustering the customers based on their reviews, and constructing the choice sets based on a sampling probability scenario that relies on product and customer clusters. The proposed scenario is called Normalized, which multiplies the product cluster and customer cluster fractions to obtain the probability sampling distribution. There are two utility functions proposed, i.e., a linear combination of product attributes only and a function that includes the interactions of product attributes and customer reviews. The methodology is implemented to a data set of laptops. The Normalized scenario performs significantly better than the baseline, Random, in predicting the test set data. Moreover, the inclusion of customer reviews into the utility function also significantly increases the predictive ability of the model. The research shows that using the product attribute data and customer reviews to construct choice sets generates choice models with higher predictive ability than randomly constructed choice sets.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.