Consumer attitudes towards purchasing refurbished smartphones in developed and developing countries: social media analytics

doi:10.21203/rs.3.rs-3123763/v1

Download PDF

Research Article

Consumer attitudes towards purchasing refurbished smartphones in developed and developing countries: social media analytics

https://doi.org/10.21203/rs.3.rs-3123763/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

In recent years, the amount of electronic waste has dramatically increased due to the short end-of-life (EOL) products, rapid expansion, and variety of electronic products; therefore, refurbishing these products is one of the practical and sustainable solutions to this issue. Customer reviews play a crucial role in determining purchasing behaviors and attitudes. To this end, a customer satisfaction model of online refurbished smartphone reviews was presented to investigate customer satisfaction dimensions (CSDs) toward refurbished smartphones by opinion mining and analyzing the data of customer reviews on Twitter from consumers of refurbished smartphones in developed and developing countries. For this purpose, we separated the refurbished reviews into two distinct categories: (i) direct evaluations of refurbished features by utilizing machine learning to identify features and determine how individuals feel about them, and (ii) indirect evaluations based on product remanufacturers' claims that refurbished and new products are comparable. The findings indicate that buyers of refurbished smartphones are most concerned with the product's features, such as its functionality, price, and guarantee conditions. In addition, we observed that while consumers in developed countries pay special attention to the environment, consumers in developing countries are more concerned with guarantees and pricing. These findings suggest economic factors as the main reason for smartphone purchases in developing countries for refurbished products. In contrast, due to greater economic prosperity in developed countries, environmental considerations were determined as the main reason for purchasing refurbished smartphones.

Refurbishment

Smartphones

Consumer behavior

Developed and developing countries

Twitter analysis

Customer satisfaction model

The idea of a circular economy (CE) is becoming increasingly popular among policymakers globally, as it is seen as a way to address the issue of excessive use of natural resources, while also offering economic advantages (Rizos and Bryhn, 2022). In a circular economy, refurbishment is a successful strategy for extending product lifespan (Wallner et al., 2022). In order to establish a closed-loop supply chain approach that promotes a sustainable circular economy, consumers' willingness is crucial when it comes to purchasing such products (Agostini et al., 2021). Furthermore, the practices of remanufacturing and refurbishing, which are both important aspects of reverse logistics, have a substantial impact on the advancement of sustainable development (Feng et al., 2021). The strategies of repairing, refurbishing, and remanufacturing aim to create a circular economy by closing the loop on product cycles and maximizing the utilization of materials and resources (Pamminger et al., 2021). Refurbishment is the process in which a specialized company collects discarded items and restores them to a functional and satisfactory condition, after which these revitalized products are made available for purchase by new customers. (van Weelden et al., 2016).

The concept of a circular economy involves reducing waste generation and minimizing the use of untapped resources, leading to more sustainable production and consumption practices. In contrast to the current linear economy where products lose their value and are discarded after use, the circular economy promotes restructuring products to achieve balance in the environment, economy, and society. The aim of a circular economy is to enhance resource productivity by creating a closed-loop system that ensures resources are kept in use for as long as possible (Ghisellini et al., 2016). Because of the widespread use of mobile devices and their short lifespan, a lot of attention has been given by researchers to the recovery and recycling rates of mobile device components. Available data shows that in 2019, there were 107 mobile phone subscriptions per 100 people, and recent data and statistics from the World Bank show that in 2022, there were 110 mobile phone subscriptions per 100 people (World Bank, 2020). The development of information technology has led to a surge in the desire for electronic devices like mobile phones. Additionally, the competitive nature of technology has decreased the lifespan of these products (Deng et al., 2017). In 2019, there was a generation of 53.6 million metric tons of e-waste items, comprising small and large equipment, temperature exchange equipment, monitors, displays, small IT and telecommunication equipment, and lamps. The quantities of each of these items were 17.4, 13.1, 10.8, 6.7, 4.7, and 0.9 million metric tons, respectively. Additionally, a global collection and recycling of all e-waste were recorded in 2019, with Europe having the highest rate of e-waste recycling, with more than 5 million tons of e-waste being recycled, while Africa had only 0.03 million (Statista, 2020). As technology continues to advance, mobile phones and smartphones are now more powerful and essential in our daily lives. As a result, more and more people are purchasing these devices and replacing them more frequently. This has led to an increase in the amount of electronic waste generated by discarded mobile phones and smartphones, making it one of the fastest-growing waste streams globally. Despite this, the collection, reuse, remanufacturing, and recycling of these devices under the Waste Electrical and Electronic Equipment (WEEE) initiative remains low (Gurita et al., 2018).

Smartphones have a substantial reusability potential, and manufacturers use relevant technology to refurbish their products (Nasiri and Shokouhyar, 2021). The bulk of the academic study devoted to extending the lifespan of smartphones has concentrated on operational elements, such as the implementation of waste management and reverse logistics processes (He et al., 2020; Santana et al., 2020). While, according to various research sources, consumer awareness and purchase behavior are among the most important impediments to electronic device reconstruction (Agostini et al., 2021; Islam et al., 2021). Moreover, even if consumers are aware of such products, they may be unaware of the refurbishment process and mistake them for used goods. Due to their association with inferior quality, refurbished items may be rejected before being considered. Other obstacles identified include a lack of availability and a desire to experiment with a new device (van Weelden et al., 2016).

The current study focuses on understanding the genuine behavior of consumers who purchase refurbished smartphones in both developed and developing countries. Additionally, the study investigates consumers' viewpoints on the factors they consider when purchasing refurbished smartphones, as well as the concerns that arise from cultural, economic, and social issues. Consequently, the study poses the following research questions:

1. What are the views of consumers in developed and developing countries when it comes to comparing refurbished and new smartphones?

2. In developed and developing countries, what are the key aspects of customer satisfaction (CSD)?

3. What is the extent of consumer satisfaction in developed and developing countries when analyzed using sentiment words?

4. What is the perception of consumers towards refurbished smartphones in both developed and developing countries?

The study proposes a framework to model online customer satisfaction in both developed and developing countries. The framework aims to answer questions regarding consumers' perceptions of refurbished smartphones compared to new phones. The study uses indirect assessments to provide comparative opinions and direct product and service evaluations to provide non-comparative opinions. Additionally, machine-based approaches are employed. The study concludes that the insights gained from this research can be used by remanufacturers and marketers to create and promote refurbished smartphones effectively, by understanding consumer attitudes and concerns in both developed and developing countries.

2.1. Using Refurbished Smartphones as a Viable Option for End-of-Life scenario

The current end-of-life (EOL) approaches can be classified into two primary categories: open-loop and closed-loop strategies (Berlin et al., 2022). End-of-life decision-making requires a comprehensive approach considering environmental, economic, sociological, business, technological, market, and legal factors when evaluating (EOL) options (Alamerew and Brissaud, 2019). One potential method for implementing circular manufacturing concepts in real-world industrial settings is through closed-loop supply chains (CLSCs) that incorporate reverse logistics. By collecting end-of-life (EOL) items from customers, reverse logistics can contribute to the supply chain's profitability, stability, and sustainability, making CLSCs a promising approach (Soleimani and Govindan, 2014). Accordingly, (Noman and Amin, 2017) focused on three provinces in Canada to investigate mobile phone reverse logistics (RL) components. Furthermore, some research has established waste management systems involving end-of-life mobile phone collection and recycling (Jang and Kim, 2010; Ongondo and Williams, 2011; Xu et al., 2016; Zhang et al., 2019). At present, manufacturers are being urged to consider adopting closed-loop supply chains (CLSCs) that ensure the retrieval of end-of-life products (EOL) for recycling or reuse, as environmental concerns continue to grow and global economic competitiveness intensifies (Delpla et al., 2021).

Refurbished products are previously owned consumer products that have been retrieved from consumers after their initial use and restored to a satisfactory condition prior to resale (Pigosso et al., 2010). In contrast, remanufacturing is the process of restoring a used product to a satisfactory condition in accordance with its original characteristics. Refurbishment is an intelligent method, especially for items that users cannot fix or will be discarded before the EOL (Wallner et al., 2020).

2.2. A comprehensive viewpoint on consumer behavior

The decision of consumers to purchase refurbished products is affected by several factors, and studies suggest that their perception of product quality plays a significant role In addition to attitude, subjective norms, and product- and price-related variables, consumer behavior toward refurbished products is complex and distinct (Pisitsankkhak and Vassanadumrongdee, 2020). Furthermore, over the past decade, researchers have sought to examine customer behavior in closed-loop supply chains (CLSCs), and the attitude of consumers has been identified as one of the most important predictors of their adoption or change in behavior toward these products (Hazen et al., 2015). When it comes to e-waste (electronic waste), researchers around the world have observed that the main categories of consumer behavior are consumption, storage, repair and reuse (R&R), disposal, and recycling (Islam et al., 2021). Moreover, the literature on consumer behavior emphasizes the significance of one's attitude when making a purchase decision (Agostini et al., 2021).

According to (Yin et al., 2014), 47.9% of customers agreed to pay to recycle their smartphones. They also investigated how Willingness to Pay (WTP) influences consumer behavior regarding mobile phone recycling. In another study, (Ananno et al., 2021) surveyed consumer behavior and e-waste management WTP in Bangladesh. According to the findings of this study, the selected WTP would cover 5–10% of recycling costs. WTP influences environmental awareness, income, and education.

2.3. Importance of consumer post-purchase behavior

The lack of customer acceptance of remanufactured products in developing and developed countries prevents closed-loop supply chains from achieving remanufacturing's maximum potential value (Wang and Hazen, 2016). One of the CLSC criteria, customers' perceptions of remanufactured products, substantially influences demand for remanufactured products (Vafadarnikjoo et al., 2018). In this context, Understanding consumer attitudes towards refurbished products and meeting their expectations within closed-loop supply chains (CLSCs) is crucial for manufacturers to maximize sales and profitability (Sharifi and Shokouhyar, 2021).

The decision-making process is crucial because users who buy products or services must go through it prior to completing the purchase, and it can also be used to evaluate customers' purchasing patterns and repeat purchases (Petcharat and Leelasantitham, 2021). Various models have been developed to represent the consumer purchasing process. ha outlined five distinct stages for purchase decisions, including need recognition, information search, alternative evaluation, purchase, and post-purchase (Kotler, 2011).

First, a study by (van Weelden et al., 2016) observed a difference in the "refurbished products purchasers" and "new products buyers" initial answers to survey questions. The first group was happy and enthusiastic about refurbished products, whereas the second group was unsure about them.

According to (Singhal et al., 2019), consumer purchase intentions are impacted favorably and substantially by attitude and subjective norms but only marginally by perceived green advantages and perceived behavior control. (Karimi et al., 2018) researched the difference between consumer behavior in purchasing decisions and consumer attributes. Furthermore, environmentally conscious consumers exhibit a preference for remanufactured goods, although they may not always view them as being eco-friendly (Abbey et al., 2015). Because refurbished smartphones are in a different set of circumstances than brand-new smartphones, pre-purchase assumptions and actual consumer behavior do not align (Nasiri and Shokouhyar, 2021).

Second, researchers have recommended the development of a more comprehensive theory of purchase intention, known as consumer behavior, to capture the discrepancy between customers' pre-purchase expectations and their behavior in the post-purchase phase of refurbished smartphones (Agostini et al., 2021). In order to provide advertising and marketing managers and manufacturers of refurbished smartphones with effective marketing strategies, we investigated the consumer behavior of refurbished smartphone users in both developed and developing countries. (Zhang and Benyoucef, 2016) suggest that companies should focus on building relationships with customers in the post-purchase stage rather than just raising awareness in the pre-purchase stage to improve their marketing strategies.

Third, in-depth interviews, online surveys, and online product reviews were utilized to collect information regarding the consumer behavior of reconstructed smartphones. In this study, we employed an online product review due to its advantages over the other two methods. One of the benefits of online product reviews is the ability to collect and analyze diverse information from a large number of consumers in the shortest amount of time at a relatively low cost (Darko and Liang, 2022). Furthermore, online product reviews provide a wealth of information about consumers' opinions, emotions, and requests, all of which are useful in determining customer satisfaction (Guo et al., 2017).

Finally, The expansion of online social networks (OSN), including blogs, Facebook, Twitter, and websites with user-generated content, has contributed to the increase of electronic word-of-mouth (eWOM) in recent years (Wang et al., 2019). One of the traits of eWOM is spontaneity, which refers to how openly consumers share their opinions on brands and describe their behavior. It is produced without the direct influence or motive of marketers and is typically driven by the desire to assist, caution, or communicate with others, and is less likely to be an untruth (Alzate et al., 2022). studies have shown that consumers' purchasing decisions are impacted by eWOM platform (Duarte et al., 2018). These platforms have evolved into potent instruments for facilitating dialogue and reaching sound judgments. In the meantime, OSN has provided a low-cost channel that can influence an unprecedented number of prospective customers, which is highly valued in the marketing industry (Wang et al., 2019).

Online social media has altered customer communication and product information sharing. When it comes to making shopping decisions, consumers are increasingly relying on the views and suggestions of social media users. In addition, social effects have a bigger influence on a consumer's social buying intention than perceived information quality (Fu et al., 2020). To fill the research gap on customer perceptions of refurbished smartphones in developed and developing countries, an exploratory study using online reviews is suitable. The study aims to investigate the post-purchase phase of consumer decision-making processes and evaluate consumer responses in the market for refurbished smartphones in both developed and developing countries.

3.1. Theoretical implication for online review mining

Customers are interested in observing the opinions of other customers via online product reviews, blogs, and social networking sites, among others. Therefore, customer feedback is crucial in determining purchasing behavior and attitude toward online shopping (Sohail et al., 2016). Customer reviews on the internet are a significant source of information for consumers evaluating items before making a purchase (Zhao et al., 2018). Customers who are considering making a purchase frequently read online product reviews given by other buyers. Such reviews give consumers a glimpse of other users' purchasing and usage experiences. Shoppers see these reviews as more reliable than seller-sponsored marketing (Thakur, 2018). Due to the pervasiveness of social media on the internet, opinion mining has become a necessary method for analyzing data, which, if successful, aids in making decisions (Sohail et al., 2016; Sun et al., 2017). Customers can use online reviews to make more informed purchasing decisions because they have read the reviews of other customers prior to making a purchase. After consumption and experience, a comparison is made between the actual outcomes and the customer's expectations. Then, feedback is provided based on performance perceptions employing sentiments as a post-purchase behavior to demonstrate purchase satisfaction.

On the other hand, compared to merchant-provided recommendations or reviews on shopping platforms, consumers prefer information provided on social networking platforms (Li and Ku, 2018; Qazi et al., 2017). Opinions are divided into two categories: comparative and non-comparative (El-Halees, 2012; Eldefrawi et al., 2019; Xu et al., 2011). Researchers determined four types of comparable opinions: non-equal gradable, equative, superlative, and non-gradable (Kansaon et al., 2020).

Alternatively, according to (Nasiri and Shokouhyar, 2021), non-comparative opinions were classified as regular opinions. They initially selected and classified eight customer satisfaction dimensions (CSDs) into three categories. "features of the product" was the most important of these categories. These features will be accessible to customers who acquire a refurbished smartphone. The literature demonstrates the significance of review mining. Generally, two types of opinions exist: comparative and non-comparative. In opinion mining, we focus on equative-type comparative opinions and regular non-comparative opinions.

Sentiment analysis, which is a component of opinion mining, is used to identify views and arguments in a given text. If there are any positive, negative, or neutral statements in comments or reviews, the algorithm attempts to uncover assertions of agreement or disagreement (Wang and Zhang, 2020). After selecting the target and utilizing the sentiment identification algorithm (Giatsoglou et al., 2017), the textual documents are classified as positive, negative, or neutral based on their polarity. For the sentiment identification stage, various strategies may be used, some of which are unsupervised and others supervised. Lexicon-based approaches are frequently used in unsupervised scenarios; lexical resources are utilized to assign polarity ratings to individual words to determine a text's overall emotion. When it comes to unsupervised scenarios, machine learning is frequently employed, with algorithms such as Nave Bayes, Neural Networks, Decision trees, Linear regression, and Support Vector Machines (SVM) used for sentiment detection tasks serving as classification problems.

3.2. Practical methods

This research study classifies online reviews of refurbished products into two groups: (i) direct evaluations that focus on specific features of the refurbished product and analyze the sentiment towards them using machine learning, and (ii) indirect evaluations that rely on the claims made by the manufacturers of refurbished products that they are comparable to new ones. In this part, we will explain the methods we used to conduct our research and gather data on customer satisfaction with refurbished smartphones in both developed and developing markets. Using the three distinct procedures depicted in Fig. 1, it will be disassembled into its component parts and described in detail. To comprehend how satisfied customers feel after making a purchase, it is crucial to collect online reviews, preprocess them, and analyze both comparative and non-comparative opinions.

To execute the steps in Fig. 1, Twitter API was used to collect data from developed and developing countries. The reviews were then preprocessed using tokenization, stop word removal, stemming, and part-of-speech (POS) tagging. Then, we separated reviews into opinions that were comparative and non-comparative. We identified comparative sentences in comparisons using entities and related keywords. Subsequently, we removed sentences that were not connected. Afterward, we extracted customer satisfaction dimensions (CSDs) using Latent Dirichlet Allocation (LDA) and word frequency statistics for non-comparative opinions. A support vector machine (SVM) determined the sentimental inclination of reviews and CSDs.

Fig .1. Post-purchase satisfaction framework for refurbished smartphones in developed and developing countries.

Step 1. Data collection

Utilizing the Twitter API to obtain data from the target countries was the first step in the data collection process. This study employed tweets from developed countries, including the United States, Canada, the United Kingdom, Australia, and Germany, and tweets from developing countries, including India, South Africa, Malaysia, Saudi Arabia, and Brazil. Twitter has become one of the most important data sources for data sharing, with most users sharing their opinions on topics such as purchasing a new device, device performance, and others. Due to the widespread use of Twitter, users' tweets will reveal concerns, challenges, or issues encountered by users of refurbished mobile phones.

Step 2. Data preprocessing

Cleaning or preprocessing data is essential before the machine learning system can prepare and highlight attributes that require processing (Drikvandi and Lawal, 2020). Many open-source NLP tools, such as the Natural Language Toolkit (NLTK), OpenNLP, and Stanford CoreNLP, support major NLP tasks (Sun et al., 2017). Python is excellent at handling text data, which has led to the development of several popular libraries for information retrieval, natural language processing, and text analytics, such as NLTK, Gensim, and spaCy (Sarkar, 2019). An extensive platform with a variety of corpora and lexical resources is the Natural Language Toolkit (NLTK). In addition, it includes the tools, interfaces, and methods required to manage and evaluate text data (Lauriola et al., 2021). Some techniques for cleaning or preprocessing text data include tokenization, stop-word removal, stemming, and part-of-speech (POS) tagging. These procedures must be implemented in our text collection (Drikvandi and Lawal, 2020). In the following sections, we will examine the strategies mentioned above are examined in detail.

"Tokenization" refers to the process of dividing a block of text into smaller units, such as words, phrases, or symbols, that are meaningful for a specific task. Although this task may seem straightforward when working with machine-readable formats, it still requires addressing certain challenges, such as removing punctuation marks and hyphens at the end of lines. Nonetheless, tokenizers can correctly process other characters, including brackets and hyphens. Additionally, tokenization can help establish the credibility of the documents being analyzed (Vijayarani and Jahani, 2016).

"Stop words removal" refers to the practice of removing words from a piece of writing that have little or no significance in order to preserve the words with the most meaning and context. Examining the frequency of solitary tokens in a large corpus of text is the most common method for identifying stop words. Stop words include words such as "a," "the," "and," and others. Every discipline and language has its own collection of stop words; therefore, there is no exhaustive list of all stop words (Sarkar, 2019). Moreover, in certain scenarios like sentiment analysis, specific words like "no" or "not" that have a negative connotation cannot be eliminated from the text since it affects the intensity of the sentiment being conveyed (Kaczmarek et al., 2021).

To fully comprehend the stemming process, we must first identify the word stems pertaining to stemming. Stems and affixes combine to form morphemes; adding affixes to word stems modifies their meaning or creates a new term; attaching affixes to word stems generates new words. This process is called inflection. Stemming is obtaining the base form of a word from its inflected form. Stemming enables various applications, including text classification, text clustering, and information retrieval, by standardizing words to their base stem regardless of their inflections. Words such as "JUMPS," "JUMPED," and "JUMPING," for instance, are created by adding affixes to the word "JUMP." This is the word stem for the base word "JUMP" (Sarkar, 2019).

"Part of speech (POS) tagging" is a process in natural language processing where words are labeled with their respective part of speech, such as nouns, verbs, adjectives, etc. It is a crucial aspect of natural language processing and serves as the basis for named entity resolution, query resolution, and word sense disambiguation (Kulkarni and Shivananda, 2021). POS extracts text features by utilizing the network's capacity for self-encoding without taking linguistic expertise into account (Xiao et al., 2022).

Step 3. Opinion mining

3.2.1 Comparative opinion mining

Comparative opinion mining is a specific area within the broader field of opinion mining. Its main purpose is to extract comparative information and relationships from text data. Comparative opinion mining focuses on identifying and analyzing comparative forms and expressions in text, which can help in understanding the relationships and opinions expressed in the text (Guo et al., 2022). Comparative opinion mining is extremely significant when attempting to assess something, as it provides a basis for comparison (Varathan et al., 2017). (Xu et al., 2011) developed a model based on mobile phone data sets. This model was used by (Nasiri and Shokouhyar, 2021) to compare the functionality of refurbished and new smartphones on Amazon and Backmarket with the following comparative relationship:

R (P1, P2, A, S)

In this research, we utilize comparative opinion mining to compare refurbished and brand-new smartphones in developed and developing countries. The products' names are represented by P1 and P2, while A denotes the characteristic name and S represents the sentimental phrase taken from customers' reviews. However, in this study, we exclude the sentimental phrases and only focus on comparing the two types of smartphones. Consumers of refurbished smartphones in both developed and developing countries believe that their functionality is identical to that of new phones. Thus, refurbished smartphones appear brand new.

Entity and keyword attributes are both required for comparative relationship extraction. In this instance, we compare two distinct products: new and refurbished. We used the keyword set in both extraction tasks as the attribute set for machine learning (ML). In this particular instance, the SVM learner produces the best results. Then, irrelevant sentences were removed to improve classification results.

3.2.2 Non-comparative opinion mining

Aspects of customer satisfaction were extracted from online reviews to model customer satisfaction. Then, sentiment orientation analysis of reviews was conducted on each customer satisfaction dimension based on the SVM learner. The process of extracting customer satisfaction dimensions (CSDs) from online reviews includes two steps, (1) extraction of CSD based on LDA and (2) sentiment orientation. Detailed descriptions of each step's particulars are provided below.

• Extraction of CSD based on LDA

This section provides a brief introduction to Latent Dirichlet Allocation (LDA). LDA is an unsupervised learning technique that does not require labeled data, making it adaptable to different languages and domains. The model automatically generates semantically related word clusters to represent each topic. Unlike many comparative opinion mining tools, the LDA model exclusively uses text data and provides more in-depth information. LDA is a reliable and interpretable approach (Guo et al., 2022). Consequently, we used LDA in our non-comparative analysis due to its reliability.

The LDA model treats each online review as a probability distribution consisting of a blend of subjects and phrases. To form a review, one must repeatedly choose a topic based on its distribution and then choose words for that topic based on the probability distribution of words associated with it (Blei et al., 2003). After conducting online reviews, we calculated the probability of each phrase using the probability formula proposed by (Nasiri and Shokouhyar, 2021):

$$p(\text{word }\mid \text{ review })=\sum _{\text{topic }} p(\text{ word }\mid \text{ topic })\times p(\text{ topic }\mid \text{ review })$$

This probability formula can be expressed as a matrix, "Review-Word" matrix is the frequency of a word in a review, the "Review-Topic" matrix is the probability of a topic in a review, and the "Topic-Word" matrix is the probability of a word in a topic. The "Review- Word" matrix is denoted as${x}_{m\times n}$, where $M$ is the number of online reviews and $N$ is the number of words in all the documents. The trained LDA model produces the "Review-Topic," "Topic-Word," and topic lists. Using the LDA generation process, the LDA model can be trained using the ${x}_{m\times n}$ acquired from the preceding formula matrix. Since we have used Twitter API to extract opinions from developed and developing countries. To evaluate the features of the refurbished smartphones themselves, we have removed all comments relating to fraud, shipping, Service and behavior of sellers. Then, in order to obtain more reliable results, we manually filtered out noisy words.

After analyzing the initial topics generated by LDA, the authors introduce a new method called SKP-LDA which incorporates sentiment word co-occurrence and knowledge pair feature extraction algorithms. They also create a word bag that takes into account the co-occurrence of emotive words, which helps to capture the emotional polarity of short texts such as microblogs. Furthermore, the SKP-LDA model utilizes topic-specific words and topic-related words as knowledge pairs that are added to the LDA model to improve its accuracy and effectiveness (Wu et al., 2021). Thus, semantic information can be derived with greater precision. Finally, we labeled each topic that can be considered for CSD for online reviews in developed and developing countries.

Our objective was to utilize SVM to identify the sentiment orientations of reviews related to each CSD. We employed the approach recommended by (Bi et al., 2019). In the set of online reviews $R$, ${R}_{i}=\left\{{r}_{i1}^{l},\dots ,{r}_{ic}^{k},\dots ,{r}_{i{C}_{i}}^{g}\right\}$denotes the set of reviews for the $i$th CSD, where ${r}_{ic}^{k}$ represents the $c$th review in ${R}_{i}$, and $R$,${C}_{i}$denotes the total number of reviews for the $i$th CSD, $l,k,g\in \{1, 2, . . . ,\text{M}\}$.To derive ${R}_{i}$, we classified the online reviews in set $R$ into sentences based on punctuations. By extracting sentences containing ${\text{w}\text{o}\text{r}\text{d}}_{ij}$in$R$, $i=\text{1,2},\dots ,I$,$j=\text{1,2},\dots ,{J}_{i}$, ${R}_{i}$ can be obtained according to ${f}_{i}=\left\{{\text{w}\text{o}\text{r}\text{d}}_{i1},{\text{w}\text{o}\text{r}\text{d}}_{i2},\dots ,{\text{w}\text{o}\text{r}\text{d}}_{iJ}\right\}.$If multiple sentences in a review relating to the same CSD, they can be consolidated.

Each ${R}_{i}$ review's sentiment is determined using SVM, a cutting-edge machine-learning technique for sentiment classification. Two phases are required for SVM sentiment classification: (I) feature construction using a bag-of-words model and (ii) training using tagged reviews. Each ${R}_{i}$ review's sentiment can be identified using the SVM sentiment classifier. The sentiment orientation of ${f}_{i}$ (the $i$th CSD) in each online review${r}_{m}$ can be calculated based on the sentiment orientation of each review in ${R}_{i}$, where$i=\text{1,2},\dots ,I$, $m=\text{1,2},\dots ,M$. As demonstrated in Table 1, we represent the ${f}_{i}$'s sentiment orientation in the online review ${r}_{m}$ where ∗ $\in \left\{\text{P}\text{o}\text{s},\text{N}\text{e}\text{g}\right\}$,$i=\text{1,2},\dots ,I$, $m=\text{1,2},\dots ,M$. The nominally coded data can be transformed into structured data using the equation below, as illustrated in Table 1. From Table 1, it is evident that if the sentiment orientation of ${f}_{i}$ in the online review ${r}_{m}$ is positive,${S}_{im}^{Pos}=1$ and ${S}_{im}^{Neg}=0$; if the sentiment orientation is negative, ${S}_{im}^{Pos}=0$ and ${S}_{im}^{Neg}=1$; and if the sentiment orientation is missing, ${S}_{im}^{Pos}=0$and ${S}_{im}^{Neg}=0$,

$$i=\text{1,2},\dots ,I, m=\text{1,2},\dots ,M.$$

$${S}_{im}^{*}=\left\{\begin{array}{c}1,\hspace{0.25em}\hspace{0.25em}\hspace{0.25em}\hspace{0.25em}\text{ }\text{i}\text{f}\text{ }\text{t}\text{h}\text{e}\text{ }\text{s}\text{e}\text{n}\text{t}\text{i}\text{m}\text{e}\text{n}\text{t}\text{ }\text{o}\text{r}\text{i}\text{e}\text{n}\text{t}\text{a}\text{t}\text{i}\text{o}\text{n}\text{ }\text{i}\text{s}\text{*}\text{ }\\ 0,\hspace{0.25em}\hspace{0.25em}\hspace{0.25em}\hspace{0.25em}\text{ }\text{o}\text{t}\text{h}\text{e}\text{r}\text{w}\text{i}\text{s}\text{e}\text{ }\end{array}i=\text{1,2},\dots ,I,m=\text{1,2},\dots ,M\right.$$

Table 1

Data on the structure of online reviews.
	CSDs
	${f}_{1}$		${f}_{2}$		$\cdots$	${f}_{7}$
Online review	${S}_{1}^{\text{P}\text{o}\text{s}}$	${S}_{1}^{\text{N}\text{e}\text{g}}$	${S}_{2}^{\text{P}\text{o}\text{s}}$	${S}_{2}^{\text{N}\text{e}\text{g}}$	$\cdots$	${S}_{7}^{\text{P}\text{o}\text{s}}$	${S}_{7}^{\text{N}\text{e}\text{g}}$
${\varvec{R}}_{1}$	0	1	0	0	$\cdots$	1	0
${\varvec{R}}_{2}$	1	0	0	0	$\cdots$	0	0
$\cdots$	$\cdots$		$\cdots$		$\cdots$	$\cdots$
${\varvec{R}}_{\varvec{M}}$	0	0	1	0	$\cdots$	0	1

In this section, a comparison is made between the results obtained in developed and developing countries. This comparison aims to analyze and highlight the commonalities and disparities in the purchasing behavior of consumers in these two types of countries. By comparing the results of developing and developed countries, we can draw a definitive and exhaustive conclusion about refurbished smartphones provided by consumers. To this end, we compiled a one-year dataset containing 29,854 online product reviews, with developed countries accounting for 16,658 reviews and developing countries for 13,196 reviews of refurbished smartphones.

4.1. Comparative appraisal result

According to product descriptions in developed countries, refurbished smartphones appear to be second-hand. In addition, the product descriptions in developing countries refer to refurbished smartphones as "used products."

According to the comparative appraisal analysis depicted in Fig. 2, 72.2% of reviews in developed countries and 68.7% in developing countries believe that the refurbished smartphones they purchased were second-hand. Thus, 27.8% of reviews in developed and 31.3% in developing countries believe that refurbished smartphones are brand new. Overall, it can be deduced that one-third of consumers of refurbished smartphones in developing and developed countries believe that a refurbished product is new (unused).

Fig .2. Comparison of refurbished and brand-new smartphones based on the frequency ratio

4.2. Non-comparative appraisal result

After tallying each review of a product's features separately for developed and developing countries, we calculated the frequency of those reviews overall. The terms "fraud," "seller's behavior and service," and "shipping" frequently appeared in customer reviews of refurbished smartphones in both developed and developing countries. However, since these terms are related to service, the purchase and purchase cycle of a product are more relevant than the product's features, we omitted them from our list of features.

In this formula, $i$ denotes the total figure of CSDs (topics), and $j$ represents the frequency with which the $i$th topic's keyword appears. Thus, we may refer to the$i$th CSD ${f}_{i}=\left\{{\text{w}\text{o}\text{r}\text{d}}_{i1},{\text{w}\text{o}\text{r}\text{d}}_{i2},\dots ,{\text{w}\text{o}\text{r}\text{d}}_{i{J}_{i}}\right\}$, where ${\text{w}\text{o}\text{r}\text{d}}_{ij}$ is the $j$th most common word in the $i$th CSD, $i=\text{1,2},\dots ,I$,$j=\text{1,2},\dots ,{J}_{i}$. After merging the sub-set topics and selecting a suitable label for each topic, we classified these CSDs into seven categories, including "Operation," "Physical characteristics," " Guarantee," "Price," "Battery health," "Environment," and "Packing." As shown in Table 2, Operation, Guarantee, and Price were the most important factors for buyers of refurbished smartphones. According to the research findings, one of the CSDs is Operation, which includes topics such as internet mobiles, camera quality, screen quality, and, in general, smartphone components' performance, hardware, and software. As can be seen, the most pressing concern is Operation in both developing countries and developed countries. After that, Price and Guarantee, which customers have discussed nearly 10,000 times and more than 8,500 times, respectively. In the remaining 4 CSDs, Physical characteristics and Battery health, have the most customer reviews in developed and developing countries.

Table 2

The LDA-based CSD topic extraction.
${\varvec{f}}_{\varvec{i}}$	CSD topic	${\varvec{J}}_{\varvec{i}}$	Total frequency in developed countries	Total frequency in developing countries	Total frequency
f1	Operation	16	8381	5849	14230
f2	Physical characteristics	13	3671	3354	7025
f3	Guarantee	10	2197	6351	8548
f4	Price	8	2989	6773	9762
f5	Battery health	7	2468	1795	4263
f6	Environment	5	1759	435	2194
f7	Packing	4	382	169	997

Table 3 shows the frequency of reviews by characteristic and the total average for developing and developed countries. We categorized extracted features as either "Main cause" or "Inciting cause." It is evident from Table 3 that consumer concerns regarding the purchase of refurbished smartphones vary between developing and developed countries. Consumers in developed countries are more concerned with inciting causes than those in developing countries. The surveys conducted on Operation, Battery health, and Physical characteristics are nearly identical in both developed and developing countries, but the surveys conducted on Guarantee and Environment are completely different; the frequency of Guarantee reviews in developing countries is three times that of developed countries, and the frequency of Environment in developed countries is also three times that of developing countries. Regarding Pricing and Packing, there is a significant disparity between reviews of refurbished smartphones from developed and developing countries, with consumers from developed countries discussing these aspects more than twice as often as those from developing countries.

Table 3

Frequency ratio and categorization of CSD topics
CSD topic	Category	Developed countries	Developing countries	Total average
Operation	Main cause	50.31%	44.32%	47.3%
Physical characteristics	Main cause	22%	25.4%	23.7%
Guarantee	Inciting cause	13.18%	48.12%	30.65%
Price	Main cause	17.94%	51.32%	34.63%
Battery health	Main cause	14.81%	13.6%	14.2%
Environment	Inciting cause	10.55%	3.29%	6.92%
Packing	Inciting cause	2.29%	1.28%	1.78%

The most significant categories and features were identified using the average frequency ratio of CSD across all online reviews, as shown in Fig. 3. Seventy-five percent of reviews identify the main cause. Regarding the Environment feature, customers who intend to purchase refurbished smartphones in developed countries have greater green knowledge than customers who intend to purchase refurbished smartphones in developing countries. They addressed this issue more than triple, which has significantly influenced their intention to purchase refurbished smartphones. In addition, the price factor is regarded as one of the primary motivations for purchasing refurbished smartphones. Since refurbished smartphones are 20–30% less expensive than new ones, consumers are willing to purchase them. According to the economic conditions of consumers in developing countries, as indicated by the surveys, nearly twice as many consumers in developed countries have discussed this issue.

Even though customer reviews can highlight a product's key features, they cannot reveal whether a customer has a positive or negative emotional disposition toward the product (Nasiri and Shokouhyar, 2021). After consulting with experts in the field, we excluded the environment, one of the obtained variables, from the category of inciting cause. Figure 4 depicts the positive and negative polarity of evaluations for each of the six quality aspects in developed and developing countries to assess potential smartphone buyers' emotional orientation. In order to assess customer satisfaction with the perceived quality of refurbished smartphones, we collected and analyzed reviews from main and inciting cause categories to determine the number of positive and negative reviews. Then, we calculated the ratio of positive and negative ratings to the total number of reviews in both developed and developing countries. Finally, we conducted reviews of customer satisfaction or dissatisfaction with each factor.

In both developed and developing countries, the ratio of positive to negative reviews for refurbished smartphones indicates customer satisfaction with these products, as shown in Fig. 4. In developed and developing countries, customer satisfaction levels for all three features — operation, physical characteristics, and battery health — are comparable. Customers are most satisfied with the operation, the most popular feature, which received the highest ratings in both developed and developing countries.

Customers have also expressed high levels of satisfaction with the physical characteristics dimension. Consumers reported a variety of operational issues, including "my phone has a hardware issue," "the camera does not work," "Wi-Fi will not connect," and "the screen suddenly went black." Most consumers who provided negative feedback regarding battery issues mentioned "the battery drains quickly" and "the battery does not look new." In developed countries, customer satisfaction with guarantees was slightly below 84%, whereas it was 67.38% in developing countries. This demonstrates that customer satisfaction with the guarantee feature is directly related to the manufacturer's performance in developed and developing countries. In developed countries, customer satisfaction with packing features is significantly high. Customers have provided positive feedback on this topic: "the packing is a stylish cardboard box without plastic" and "the packing is intact and eco-friendly." Comparatively, customer satisfaction with packing in developing countries was lower at 71.34%. In developing countries, 81.43% of customers were satisfied with the prices they paid. In contrast, in developed countries, this under at 68.72%. This suggests that economic factors influence customer satisfaction in developing countries.

As indicated in Fig. 5, the highest levels of customer satisfaction in developed countries were associated with packing (88.27%), operation (85.48%), and guarantee (83.69%), respectively. Moreover, in developing countries, operation (88.73%), price (81.41%), and physical characteristics (74.22%) had the highest levels of consumer satisfaction. The findings of this study indicate that in both developing and developed countries, consumer satisfaction with battery health, physical characteristics, and health is nearly identical. Consumers in developed countries were much more satisfied with perceived guarantee and packing than those in developing countries, where satisfaction was primarily based on price.

The shopping experience provides value and is defined as the consumer's interaction with an item (e.g., a product or service). Therefore, value, sometimes referred to in the literature as perceived value, is subjective in nature. According to the consumer, perceived value is the overall assessment of a product or service (Lu et al., 2022). (Wang and Hazen, 2016) investigated how consumers perceive remanufactured products' costs, quality, and environmental benefits. Customers' perceptions of value were positively affected by the quality, cost, and environmental benefits associated with these products, while their perceptions of risk were negatively affected by the quality and cost benefits.

According to the study's findings, the perception of incentives in developed and developing countries is influenced by product price, customer knowledge of the green concept, smartphone performance, and quality. Previous research has shown that consumers are more likely to purchase refurbished products due to environmental consciousness, economic benefits, and quality benefits (Nasiri and Shokouhyar, 2021; Wang and Hazen, 2016). Furthermore, (Nasiri and Shokouhyar, 2021) separated perceived values into personal and social dimensions. This article investigated these two aspects in developing and developed countries.

Due to consumers' low-income levels in developing countries, it is crucial to consider a budget when purchasing refurbished smartphones. In these countries, due to economic factors, people's attitudes are more dependent on the provision of basic needs, as they pay less attention to high-level issues such as environmental concerns and sustainable development until their fundamental needs are met. According to (Mugge et al., 2017), the desire of consumers in developing countries to purchase refurbished smartphones is compounded by the similarity of these devices to new ones in appearance and performance.

In both developed and developing countries, a significant proportion of consumers consider the similarity between the physical characteristics and operationally of smartphones. In addition, providing a guarantee for refurbished smartphones is another benefit that encourages consumers to purchase these devices. In this context, the economic and functional aspects of refurbished smartphones are also considered, assuring the consumer that the money spent on these devices will yield a financial return. Additionally, it is guaranteed to solve consumers' technical issues from an operational standpoint. As previously stated, economic factors play a crucial role in the lives of people in developing countries. Therefore, in these countries, governments have prioritized and made greater efforts to address the economic issues of the populace (Shahrasbi et al., 2021). Moreover, environmental issues and related concerns are not among the government's top priorities. This reduces funding for addressing environmental issues and fostering individual and social consciousness.

The economic discussion surrounding the purchase of refurbished smartphones is also potentially important in developed countries. (Sharifi and Shokouhyar, 2021) demonstrated that financial incentives encourage consumers to use these phones. However, in developed countries, the people's basic needs have been met, their economic problems have been partially resolved, and relative prosperity has been created in society (Shahrasbi et al., 2021). As a result, individuals focus on higher needs, such as social issues, and are more concerned with environmental concerns. (Chen et al., 2019) also reported that the greater the environmental consciousness of a consumer, the greater his desire for refurbished products. According to (Gaur et al., 2015), consumers who are environmentally conscious or "green concept" are more likely to purchase remanufactured products due to their strong values and environmental consciousness.

(Michaud and Llerena, 2010) state that refurbished products are green products, and (Nasiri and Shokouhyar, 2021) believe that refurbished smartphones are also considered green products due to their environmental benefits. In discussing the social dimension, it can be asserted that in developed countries, in addition to providing social welfare and resolving economic issues, governments advance their plans and programs toward sustainable development on a higher level. (Gaur et al., 2015) found that social norms have successfully advocated for and encouraged consumers to utilize refurbished products. He also emphasized the importance of advertising to increase consumer willingness to purchase refurbished products. The government and retailers must work together to promote remanufactured products as "as good as new" because quality and price are the most important factors and customers cannot verify quality before using the product (Alyahya et al., 2023). Over the past several decades, text has been produced at a never-before-seen rate. One of the primary reasons for this is the digitization of information, which makes data transfer cheaper and faster than traditional methods (Wagh et al., 2021). As a result of their open structure and access to vast data samples, online text reviews reflect consumer opinions more inclusively and exhaustively. Consumers attempt to express their opinions about the products and services they use on online social networks, which serve as valuable sources of information for businesses, governments, and professionals.

Using Twitter social networks, data mining, and sentiment analysis, we investigated the opinions of consumers in developed and developing countries regarding refurbished smartphones. Consequently, we first used data collected from tweets in developed and developing countries to determine consumer behavior after using refurbished smartphones. Following preprocessing with natural language processing tasks, we categorized the opinions as comparative or non-comparative. Using comparable search terms in comparative reviews permits us to evaluate refurbished and brand-new smartphone devices. As part of our analysis of non-comparative perspectives, seven customer satisfaction dimensions (CSDs) were initially extracted and divided into two groups: inciting cause and Main cause. We examined these two categories in both developed and developing countries. This research assists in comprehending the perceived value of purchasing refurbished smartphones in developed and developing countries. We observed that customers in developed and developing countries were satisfied with the devices regarding the stated features.

In addition, we examined the perceived values in two personal and social dimensions in these countries. This study's findings indicate that economic factors are one of the determining factors in developing countries in the individual dimension. Due to these countries' lack of economic prosperity, consumers are price-sensitive and therefore encouraged to purchase refurbished smartphones. According to this study, the after-sales service, particularly the "guarantee," is one of the other major advantages of purchasing smartphones. In contrast, studies indicate that economic prosperity has prompted consumers in developed countries to pay attention to environmental issues; even though consumers in these countries also consider the economic factor, their green consciousness motivates them to purchase refurbished smartphones.

Funding Statement

I, Milad Mounesan, hereby certify that the paper "Consumer attitudes towards purchasing refurbished smartphones in developed and developing countries: social media analytics" was written by me. The authors did not receive support from any organization for the submitted work. No funding was received to assist with the preparation of this manuscript.

Ethical Statement

Hereby, I Milad Mounesan consciously assure that for the manuscript “Consumer attitudes towards purchasing refurbished smartphones in developed and developing countries: social media analytics” the following is fulfilled:

1) This material is the authors' own original work, which has not been previously published elsewhere.

2) The paper is not currently being considered for publication elsewhere.

3) The paper reflects the authors' own research and analysis in a truthful and complete manner.

4) The paper properly credits the meaningful contributions of co-authors and co-researchers.

5) The results are appropriately placed in the context of prior and existing research.

6) All sources used are properly disclosed (correct citation). Literally copying of text must be indicated as such by using quotation marks and giving proper reference.

Conflict of Interest

To:

Dear Mr. Antonio Di Nola and Raffaele Cerulli

University of Salerno, Italy

I am submitting a manuscript of consideration for publication in the Journal of Soft computing. The manuscript is entitled “Consumer attitudes towards purchasing refurbished smartphones in developed and developing countries: social media analytics”.

I confirm that this manuscript has not been published elsewhere and is not under consideration by another journal.

A Customer Satisfaction Model of Refurbished Smartphones from Developed and Developing Countries to Investigate Customer Satisfaction Dimensions (CSD) of Refurbished Smartphones Using Twitter Data as the main data corpus with Machine Learning Techniques Provided.

Consumer satisfaction is really important. This paper assists businesses in analyzing post-purchase consumer behavior and its effectiveness. The post-purchase behaviors of consumers are frequently essential for organizations to assist managers in better guiding the perceived value of consumers.

Thank you very much for your consideration.

I certify that I have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript.

Yours Sincerely,

Milad Mounesan

Master's student in information technology management atShahid Beheshti University of Tehran,

Tehran, Iran.

Tel: +98-937-8489055

E-mail: [email protected]

Abbey JD, Meloy MG, Blackburn JD, Guide VDR (2015) Consumer Markets for Remanufactured and Refurbished Products. Calif Manage Rev 57:26–42
Agostini L, Bigliardi B, Filippelli S, Galati F (2021) Seller reputation, distribution and intention to purchase refurbished products. J Clean Prod 316:128296
Alamerew YA, Brissaud D (2019) Circular economy assessment tool for end of life product recovery strategies. J Remanufacturing 9:169–185. https://doi.org/10.1007/s13243-018-0064-8
Alyahya M, Agag G, Aliedan M, Abdelmoety ZH, Daher MM (2023) A sustainable step forward: Understanding factors affecting customers’ behaviour to purchase remanufactured products. J Retail Consum Serv 70:103172. https://doi.org/10.1016/j.jretconser.2022.103172
Alzate M, Arce-Urriza M, Cebollada J (2022) Mining the text of online consumer reviews to analyze brand image and brand positioning. J Retail Consum Serv 67:102989. https://doi.org/10.1016/j.jretconser.2022.102989
Ananno AA, Masud MH, Dabnichki P, Mahjabeen M, Chowdhury S (2021) Survey and analysis of consumers’ behaviour for electronic waste management in Bangladesh. J Environ Manage 282:111943
Berlin D, Feldmann A, Nuur C (2022) The relatedness of open- and closed-loop supply chains in the context of the circular economy; Framing a continuum. Clean Logist Supply Chain 4:100048. https://doi.org/10.1016/j.clscn.2022.100048
Bi J, Liu Y, Fan Z-P, Cambria E (2019) Modelling customer satisfaction from online reviews using ensemble neural network and effect-based Kano model. Int J Prod Res 57:7068–7088
Blei DM, Ng A, Jordan MI (2003) Latent Dirichlet Allocation J Mach Learn Res 3:993–1022
Chen Y, Wang J, Yu Y (2019) A Study on Consumers’ Willingness to Pay for Remanufactured Products: A Study Based on Hierarchical Regression Method. Front Psychol 10:1–9. https://doi.org/10.3389/fpsyg.2019.02044
Darko AP, Liang D (2022) Modeling customer satisfaction through online reviews: A FlowSort group decision model under probabilistic linguistic settings. Expert Syst Appl 195:116649
Delpla V, Kenné J-P, Hof LA (2021) Circular manufacturing 4.0: towards internet of things embedded closed-loop supply chains. Int. J. Adv. Manuf. Technol
Deng W-J, Giesy JP, So CS, Zheng H-L (2017) End-of-life (EoL) mobile phone management in Hong Kong households. J Environ Manage 200:22–28
Drikvandi R, Lawal OO (2020) Sparse Principal Component Analysis for Natural Language Processing. Ann Data Sci 1–17
Duarte P, Costa e Silva S, Ferreira MB (2018) How convenient is it? Delivering online shopping convenience to enhance customer satisfaction and encourage e-WOM. J Retail Consum Serv 44:161–169. https://doi.org/10.1016/j.jretconser.2018.06.007
El-Halees A (2012) Opinion Mining from Arabic Comparative Sentences, in: 13th International Arab Conference on Information Technology ACIT ASSESSMENT. pp. 265–271
Eldefrawi MM, Elzanfaly DS, Farhan MS, Eldin AS (2019) Sentiment analysis of Arabic comparative opinions. SN Appl Sci 1:1–11
Feng D, Shen C, Pei Z (2021) Production decisions of a closed-loop supply chain considering remanufacturing and refurbishing under government subsidy. Sustain. Prod. Consum
Fu JR, Lu IW, Chen JHF, Farn CK (2020) Investigating consumers’ online social shopping intention: An information processing perspective. Int J Inf Manage 54. https://doi.org/10.1016/j.ijinfomgt.2020.102189
Gaur J, Amini M, Banerjee P, Gupta R (2015) Drivers of consumer purchase intentions for remanufactured products a study of Indian consumers relocated to the USA. Qual. Mark Res 18:30–47. https://doi.org/10.1108/QMR-01-2014-0001
Ghisellini P, Cialani C, Ulgiati S (2016) A review on circular economy: the expected transition to a balanced interplay of environmental and economic systems. J Clean Prod 114:11–32
Giatsoglou M, Vozalis MG, Diamantaras KI, Vakali A, Sarigiannidis G, Chatzisavvas KC (2017) Sentiment analysis leveraging emotions and word embeddings. Expert Syst Appl 69:214–224
Guo Y, Barnes SJ, Jia Q (2017) Mining meaning from online ratings and reviews: Tourist satisfaction analysis using latent dirichlet allocation. Tour Manag 59:467–483
Guo Y, Wang F, Xing C, Lu X (2022) Mining multi-brand characteristics from online reviews for competitive analysis: A brand joint model using latent Dirichlet allocation. Electron Commer Res Appl 53:101141. https://doi.org/10.1016/j.elerap.2022.101141
Gurita N, Fröhling M, Bongaerts JC (2018) Assessing potentials for mobile/smartphone reuse/remanufacture and recycling in Germany for a closed loop of secondary precious and critical metals. J Remanufacturing 8:1–22
Hazen BT, Mollenkopf DA, Wang Y (2015) Remanufacturing for the Circular Economy: An Examination of Consumer Switching Behavior. Bus Strateg Environ 26:451–464
He P, Hu G, Wang C, Hewage KN, Sadiq R, Feng H (2020) Analyzing present and future availability of critical high-tech minerals in waste cellphones: A case study of India. Waste Manag 119:275–284
Islam T, Huda N, Baumber A, Shumon RH, Zaman A, Ali F, Hossain R, Sahajwalla V (2021) A global review of consumer behavior towards e-waste and implications for the circular economy. J Clean Prod 316:128297
Jang YC, Kim M (2010) Management of used \& end-of-life mobile phones in Korea: A review. Resour Conserv Recycl 55:11–19
Kaczmarek I, Iwaniak A, Świetlicka A, Piwowarczyk M, Nadolny A (2021) A machine learning approach for integration of spatial development plans based on natural language processing. Sustain. Cities Soc
Kansaon D, Brandão MA, Reis JCS, Barbosa M, de Sousa Matos B, Benevenuto F (2020) Mining Portuguese Comparative Sentences in Online Reviews. Proc. Brazilian Symp. Multimed. Web
Karimi S, Holland CP, Papamichail KN (2018) The impact of consumer archetypes on online purchase decision-making processes and outcomes: A behavioural process perspective. J. Bus. Res
Kotler P (2011) Marketing management, millenium edition
Kulkarni A, Shivananda A (2021) Natural Language Processing Recipes: Unlocking Text Data with Machine Learning and Deep Learning Using Python. Nat. Lang. Process. Recipes
Lauriola I, Lavelli A, Aiolli F (2021) An introduction to Deep Learning in Natural Language Processing: Models, techniques, and tools. Neurocomputing
Li C-Y, Ku Y-C (2018) The power of a thumbs-up: Will e-commerce switch to social commerce? Inf. Manag 55:340–357
Lu B, Yan L, Chen Z (2022) Perceived values, platform attachment and repurchase intention in on-demand service platforms: A cognition-affection-conation perspective. J. Retail. Consum. Serv
Michaud C, Llerena D (2010) Green consumer behaviour: an experimental analysis of willingness to pay for remanufactured products. Bus Strateg Environ 20:408–420
Mugge R, Jockin BGW, Bocken NMP (2017) How to sell refurbished smartphones? An investigation of different customer groups and appropriate incentives. J Clean Prod 147:284–296
Nasiri MS, Shokouhyar S (2021) Actual consumers’ response to purchase refurbished smartphones: Exploring perceived value from product reviews in online retailing. J Retail Consum Serv 62:102652. https://doi.org/10.1016/j.jretconser.2021.102652
Noman R, Amin SH (2017) Characteristics of cellphones reverse logistics in Canada. J Remanufacturing 7:181–198
Ongondo FO, Williams ID (2011) Greening academia: use and disposal of mobile phones among university students. Waste Manag 31 7:1617–1634
Pamminger R, Glaser S, Wimmer W (2021) Modelling of different circular end-of-use scenarios for smartphones. Int J Life Cycle Assess 26:470–482
Petcharat T, Leelasantitham A (2021) A retentive consumer behavior assessment model of the online purchase decision-making process. Heliyon 7. https://doi.org/10.1016/j.heliyon.2021.e08169
Pigosso DCA, Zanette ET, Filho AG, Ometto AR, Rozenfeld H (2010) Ecodesign methods focused on remanufacturing. J Clean Prod 18:21–31
Pisitsankkhakarn R, Vassanadumrongdee S (2020) Enhancing purchase intention in circular economy: An empirical evidence of remanufactured automotive product in Thailand. Resour Conserv Recycl 156:104702
Qazi A, Tamjidyamcholo A, Raj RG, Hardaker G, Standing C (2017) Assessing consumers’ satisfaction and expectations through online opinions: Expectation and disconfirmation approach. Comput Hum Behav 75:450–460
Rizos V, Bryhn J (2022) Implementation of circular economy approaches in the electrical and electronic equipment (EEE) sector: Barriers, enablers and policy insights. J. Clean. Prod
S V, R, J (2016) Text Mining: open Source Tokenization Tools – An Analysis. Adv Comput Intell An Int J 3:37–47. https://doi.org/10.5121/acii.2016.3104
Santana JCC, Guerhardt F, Franzini CE, Ho LL, Júnior SERR, Cânovas G, Yamamura CLK, Vanalle RM, Berssaneti FT (2020) Refurbishing and recycling of cell phones as a sustainable process of reverse logistics: A case study in Brazil. J Clean Prod 124585
Sarkar D (2019) Text Analytics with Python - A Practitioner’s Guide to Natural Language Processing. in: Text Analytics with Python
Shahrasbi A, Shokouhyar S, Zeidyahyaee N (2021) Consumers’ behavior towards electronic wastes from a sustainable development point of view: An exploration of differences between developed and developing countries. Sustain Prod Consum 28:1736–1756. https://doi.org/10.1016/j.spc.2021.09.016
Sharifi Z, Shokouhyar S (2021) Promoting consumer’s attitude toward refurbished mobile phones: A social media analytics approach. Resour Conserv Recycl 167:105398. https://doi.org/10.1016/j.resconrec.2021.105398
Singhal D, Jena SK, Tripathy S (2019) Factors influencing the purchase intention of consumers towards remanufactured products: a systematic review and meta-analysis. Int J Prod Res 57:7289–7299
Sohail SS, Siddiqui J, Ali R (2016) Feature extraction and analysis of online reviews for the recommendation of books using opinion mining technique. Perspect Sci 8:754–756
Soleimani H, Govindan K (2014) Reverse logistics network design and planning utilizing conditional value at risk. Eur J Oper Res 237:487–497
Statista Electronic waste generated worldwide in 2019 (in million metric tons) available at: https://www.statista.com/study/56028/global-electronic-waste/ [WWW Document], n.d. URL https://www.statista.com/study/56028/global-electronic-waste/ (accessed 9.17.22)
Sun S, Luo C, Chen J (2017) A review of natural language processing techniques for opinion mining systems. Inf Fusion 36:10–25
Thakur R (2018) Customer engagement and online reviews. J Retail Consum Serv 41:48–59. https://doi.org/10.1016/j.jretconser.2017.11.002
Vafadarnikjoo A, Mishra N, Govindan K, Chalvatzis K (2018) Assessment of consumers’ motivations to purchase a remanufactured product by applying Fuzzy Delphi method and single valued neutrosophic sets. J Clean Prod 196:230–244. https://doi.org/10.1016/j.jclepro.2018.06.037
van Weelden E, Mugge R, Bakker C (2016) Paving the way towards circular consumption: exploring consumer acceptance of refurbished mobile phones in the Dutch market. J Clean Prod 113:743–754
Varathan KD, Giahanou A, Crestani FA (2017) Comparative opinion mining: A review. J Assoc Inf Sci Technol 68
Wagh V, Khandve S, Joshi I, Wani A, Kale GV, Joshi R (2021) Comparative Study of Long Document Classification. TENCON 2021–2021 IEEE Reg. 10 Conf. 732–737
Wallner T, Magnier L, Mugge R (2022) Do consumers mind contamination by previous users? A choice-based conjoint analysis to explore strategies that improve consumers’ choice for refurbished products. Resour. Conserv. Recycl
Wallner T, Magnier L, Mugge R (2020) An Exploration of the Value of Timeless Design Styles. for the Consumer Acceptance of Refurbished Products. Sustainability
Wang J, Wang X, Li Y (2019) A discrete electronic word-of-mouth propagation model and its application in online social networks. Phys. A Stat. Mech. its Appl
Wang K, Zhang Y (2020) Topic Sentiment Analysis in Online Learning Community from College Students. J Data Inf Sci 5:33–61
Wang Y, Hazen BT (2016) Consumer product knowledge and intention to purchase remanufactured products. Int J Prod Econ 181:460–469
World bank open data Mobile cellular subscriptions (per 100 people) from 1960 to 2020 years | ITU [WWW Document], n.d. URL https://data.worldbank.org/indicator/IT.CEL.SETS.P2 (accessed 9.17.22)
Wu D, Yang R, Shen C (2021) Sentiment word co-occurrence and knowledge pair feature extraction based LDA short text clustering algorithm. J Intell Inf Syst 56:1–23. https://doi.org/10.1007/s10844-020-00597-7
Xiao Y, Li C, Thürer M, Liu Y, Qu T (2022) User preference mining based on fine-grained sentiment analysis. J Retail Consum Serv 68:103013. https://doi.org/10.1016/j.jretconser.2022.103013
Xu C, Zhang W, He W, Li G, Huang J (2016) The situation of waste mobile phone management in developed countries and development status in China. Waste Manag 58:341–347
Xu K, Liao SS, Li J, Song Y (2011) Mining comparative opinions from customer reviews for Competitive Intelligence. Decis Support Syst 50:743–754
Yin J, Gao Y, Xu H (2014) Survey and analysis of consumers’ behaviour of waste mobile phone recycling in China. J Clean Prod 65:517–525
Zhang KZK, Benyoucef M (2016) Consumer behavior in social commerce: A literature review. Decis Support Syst 86:95–108. https://doi.org/10.1016/j.dss.2016.04.001
Zhang Y, Qu Y, Wang W, Yu SL, Liu Y (2019) Joint collection mode of waste mobile phones based on residents’ preferences: A case of Dalian in China. J. Clean. Prod
Zhao K, Stylianou AC, Zheng Y (2018) Sources and impacts of social influence from online anonymous user reviews. Inf Manag 55:16–30

Download PDF

Version 1

posted

You are reading this latest preprint version

	CSDs
	\({f}_{1}\)		\({f}_{2}\)		\(\cdots\)	\({f}_{7}\)
Online review	\({S}_{1}^{\text{P}\text{o}\text{s}}\)	\({S}_{1}^{\text{N}\text{e}\text{g}}\)	\({S}_{2}^{\text{P}\text{o}\text{s}}\)	\({S}_{2}^{\text{N}\text{e}\text{g}}\)	\(\cdots\)	\({S}_{7}^{\text{P}\text{o}\text{s}}\)	\({S}_{7}^{\text{N}\text{e}\text{g}}\)
\({\varvec{R}}_{1}\)	0	1	0	0	\(\cdots\)	1	0
\({\varvec{R}}_{2}\)	1	0	0	0	\(\cdots\)	0	0
\(\cdots\)	\(\cdots\)		\(\cdots\)		\(\cdots\)	\(\cdots\)
\({\varvec{R}}_{\varvec{M}}\)	0	0	1	0	\(\cdots\)	0	1

Consumer attitudes towards purchasing refurbished smartphones in developed and developing countries: social media analytics

Status:

Version 1

Abstract

Figures

1. Introduction

2. Literature review

2.1. Using Refurbished Smartphones as a Viable Option for End-of-Life scenario

2.2. A comprehensive viewpoint on consumer behavior

2.3. Importance of consumer post-purchase behavior

3. Research methodology

3.1. Theoretical implication for online review mining

3.2. Practical methods

3.2.1 Comparative opinion mining

3.2.2 Non-comparative opinion mining

4. Results

4.1. Comparative appraisal result

4.2. Non-comparative appraisal result

5. Discussion and conclusion

Declarations

References

Status:

Version 1