Table 2 - uploaded by Ana Isabel Canhoto
Content may be subject to copyright.
Example of disparity in emotional state classification.

Example of disparity in emotional state classification.

Source publication
Article
Full-text available
The ever-growing volume of brand-related conversations on social media platforms has captivated the attention of academics and practitioners, as the analysis of those conversations promises to offer unparalleled insight into consumers’ emotions. This article takes a step back from the hype, and investigates the vulnerabilities related to the analys...

Contexts in source publication

Context 1
... tended to occur around the positive emotion 'joy'. In turn, differences were particularly marked for the emotion 'surprise' -this was evident both between manual analysis vs. automated, and between the various automated software, as illustrated by the quote provided in Table 2. ...
Context 2
... you #Nespresso' (entry 28). Similarly, emotions that were clearly positive, like 'joy', showed higher rates of agreement than those that were neutral or negative (Table 2). By contrast, an example of a problematic sentence is the entry 'Think I need an IV of caffeine today. ...

Similar publications

Article
Full-text available
Given the lack of Arabic dialect text corpora in comparison with what is available for dialects of English and other languages, there is a need to create dialect text corpora for use in Arabic natural language processing. What is more, there is an increasing use of Arabic dialects in social media, so this text is now considered quite appropriate as...

Citations

... Sentiment analysis research has become popular over the past two decades [40,61,62]; as more efficient sentiment classification models are devised [63] and studies have compared automated analysis of conversations on social media with manual approaches [64]. ...
... In other studies, Microsoft Azure has been found to yield better results when compared to other analyser tools such as Stanford NLP [64], IBM Watson Natural Language Understanding, OpinionFinder 2.0 and Sentistrength [70]. However, as Azure only identifies polarity, it is a less accurate method of measuring an individual's opinion towards a topic compared to other approaches such as VADER [71] and so part of this study compared the sentiment analysis approaches of Microsoft Azure and VADER. ...
Article
Full-text available
Vaccine hesitancy is an ongoing concern, presenting a major threat to global health. SARS-CoV-2 COVID-19 vaccinations are no exception as misinformation began to circulate on social media early in their development. Twitter’s Application Programming Interface (API) for Python was used to collect 137,781 tweets between 1 July 2021 and 21 July 2021 using 43 search terms relating to COVID-19 vaccines. Tweets were analysed for sentiment using Microsoft Azure (a machine learning approach) and the VADER sentiment analysis model (a lexicon-based approach), where the Natural Language Processing Toolkit (NLTK) assessed whether tweets represented positive, negative or neutral opinions. The majority of tweets were found to be negative in sentiment (53,899), followed by positive (53,071) and neutral (30,811). The negative tweets displayed a higher intensity of sentiment than positive tweets. A questionnaire was distributed and analysis found that individuals with full vaccination histories were less concerned about receiving and were more likely to accept the vaccine. Overall, we determined that this sentiment-based approach is useful to establish levels of vaccine hesitancy in the general public and, alongside the questionnaire, suggests strategies to combat specific concerns and misinformation.
... Mechanized supposition investigation systems are regularly employed to arrange text-based records into predefined categories reflecting the extremity of conclusion referred to within the content. As of late, [14] have tried a relative study of computerized and manual social media discussions. Their findings demonstrate low understanding degrees in manual and mechanized investigations, which has "grave concern given the notoriety of the last in shopper explore" [14]. ...
... As of late, [14] have tried a relative study of computerized and manual social media discussions. Their findings demonstrate low understanding degrees in manual and mechanized investigations, which has "grave concern given the notoriety of the last in shopper explore" [14]. There are a number of reasons for the mechanized characterization of communicated slant within social media discussions. ...
... In addition, notion examination encounters problems in the utilization of natural language processing (NLP) on non-structured content, CGC, and an average of online life discussions. For instance, CGC content typically reflects the casual nature and moment of correspondence through web-based networking media [14]. The substance is normally a free-streaming content and easily goes in its word and sentence structure use [94],generally incorporates shortened forms, emoticons, emojis, incorrect spellings, and makes use of an SMS-like linguistic structure on a regular basis, which is not satisfactorily bolstered by current slant investigation strategies. ...
Preprint
Full-text available
The volume of discussions concerning brands within social media provides digital marketers with great opportunities for tracking and analyzing the feelings and views of consumers toward brands, products, influencers, services, and ad campaigns in CGC. The present study aims to assess and compare the performance of firms and celebrities (i.e., influencers that with the experience of being in an ad campaign of those companies) with the automated sentiment analysis that was employed for CGC at social media while exploring the feeling of the consumers toward them to observe which influencer (of two for each company) had a closer effect with the corresponding corporation on consumer minds. For this purpose, several consumer tweets from the pages of brands and influencers were utilized to make a comparison of machine learning and lexicon-based approaches to the sentiment analysis through the Naive algorithm (lexicon-based) and Naive Bayes algorithm (machine learning method) and obtain the desired results to assess the campaigns. The findings suggested that the approaches were dissimilar in terms of accuracy; the machine learning method yielded higher accuracy. Finally, the results showed which influencer was more appropriate according to their existence in previous campaigns and helped choose the right influencer in the future for our company and have a better, more appropriate, and more efficient ad campaign subsequently. It is required to conduct further studies on the accuracy improvement of the sentiment classification. This approach should be employed for other social media CGC types. The results revealed decision-making for which sentiment analysis methods are the best approaches for the analysis of social media. It was also found that companies should be aware of their consumers' sentiments and choose the right person every time they think of a campaign.
... Nowadays, social media (SM) platforms have become popular among users as a source of valuable information regarding sales ( Lau et al., 2018 ;Jeong et al., 2019 ;Rathore and Ilavarasan, 2020 ), marketing ( Guerreiro and Moro, 2017 ;Moro et al., 2018 ;, hotel industry ( Calheiros et al., 2017 ;Geetha et al., 2017 ;Xu et al., 2017 ), hospitality ( Kim and Im, 2018 ), airline industry Tian et al., 2019 ;Punel and Ermagun, 2018 ;Martin-Domingo et al., 2019 ), entertainment , business ( Canhoto and Padmanabhan, 2015 ;Abrahams et al., 2015 ;Fan et al., 2013 ;Lee, 2018 ;He et al., 2019 ;Mingione et al., 2020 ;Greco and Polli, 2020 ;Liu, 2020 ), tourism Ainin et al., 2020 ), fraud detection , water business ( Pawsey et al., 2018 ) Before going to the actual purchase of products or services, customers or consumers utilize the information present on the social media platforms (e.g. Twitter and Facebook) to make purchase decisions ( Li et al., 2018 ). ...
... In stock performance, negative sentiments are more impactful as compared to pos-itive sentiments of customers or consumers ( Liu, 2020 ). Emotional text mining ( Canhoto and Padmanabhan, 2015 ) also applies to big data UGC to identify valuable customers who share similar brands on social media ( Mingione et al., 2020 ;Greco and Polli, 2020 ). In the business prospectus, customers' knowledge plays an essential role in the enhancement of the products and services of the organizations . ...
Article
Full-text available
The importance of text mining is increasing in services management as the access to big data is increasing across digital platforms enabling such services. This study adopts a systematic literature review on the application of text mining in services management. First, we analyzed the literature on which has used text mining methods like Sentiment Analysis, Topic Modeling, and Natural language Processing (NLP) in reputed business management journals. Further, we applied visualization tools for text mining and the topic association to understand the dominant themes and relationships. The analysis highlighted that social media analysis, market analysis, competitive intelligence are the most dominant themes while other themes like risk management and fake content detection are also explored. Further, based on the analysis, future research agenda in the field of text mining in services management has been indicated.
... In short, prior work on marketing has made moral social media storms a major advertising phenomenon, but the shades that occur in these moral events have not yet been studied in detail. Some research has shown that businesses are well-advised to seek more assertive responses, particularly where brand supporters can protect their brand [57]. ...
Article
Full-text available
In the past recent years, WhatsApp and WeChat have surprisingly fast growth. Facebook as well became the first social network to reach 1 billion active users every month. The presence of social media is an expectation for brands instead of an exception to the rule. Social events and shared information within your target market will help you understand developments in the industry. The opportunity to expose patterns in business in real-time is a potential business intelligence goldmine. The worldwide rate of social penetration reached 49% in 2020, with the highest penetration rates in East Asia and North America. Instagram enables users, through their standards of credibility, authenticity, and transparency, to develop themselves. Influencers from social media have a personal recognizable identity, also known as the "true brand" An influencer has tools and values that can motivate many other followers to increase their presence in the media. Even if these leads do not directly buy via social, awareness-raising can lead them to become full-time buyers. The overwhelming majority of users on Instagram are under the age of 30 according to recent Social Media demographics. Marketers face a dilemma: increased people want businesses to take a social stand, but 79% of CMOs fear that their capacity to attract consumers will be adversely affected. Businesses can mitigate negative emotions by providing positive information to popular social media users. Marketing managers will encourage consumers through tournaments and influencer programmers to engage in contact practices so customers can evangelize and encourage their loyalty to the organization through the creation and delivery of user-generated content.
... From a listening perspective, current work on identifying topics and trends in usergenerated content is still crude. There are often substantial gaps between automated and manual analyses (Canhoto and Padmanabhan 2015). Future research can improve the effectiveness of listening by pursuing three directions. ...
Article
Full-text available
Over the past two decades, everyday users have become a prominent force in the advertising landscape. They actively participate in conversations with and about brands by creating, amplifying, and interacting with brand-related messages. These user activities generate large volumes of structured and unstructured data that advertisers can mine to understand consumer interests and preferences. In this article, we survey insights from the user-generated content literature through the computational advertising lens to offer a road map for future research. Specifically, we discuss three roles that users play—as creators, metavoicers, and propagators. For each role, we present key research areas that can benefit from a computational approach, identify the opportunities and challenges, and propose questions for future research. We also discuss the practical implications of applying computational methods to study users and user-generated content for advertisers.
... There are often substantial gaps between automated and manual analyses (Canhoto and Padmanabhan 2015). Future research can improve the effectiveness of listening by pursuing three directions. ...
... By analyzing UGC, companies can gain valuable information about why customers like or dislike a company's products and services (Li and Li, 2013). Opinions published on social media can represent authentic information voiced directly by consumers (Presi et al., 2014), as they tend to be produced without external pressure (Canhoto and Padmanabhan, 2015). Therefore, by studying UGC, different results can be obtained more readily than with traditional survey research. ...
... Moreover, one part of the text can express more than one sentiment and refer to more than one object, which creates uncertainty for sentiment classification. Finally, critical reviews of brands often also have irony and sarcasm that is difficult to identify, which makes it challenging to classify sentiments (Canhoto and Padmanabhan, 2015). We considered the first issue by instructing the Table V. ...
... This comment would fit both Quality(+) and Pleasantness(+) classes, as the word "best" represents by the vocabulary, quality and the word "friendly" agreeableness. In such situations, the classification decision could be based on the overall picture that is generated from the comment (Canhoto and Padmanabhan, 2015). In the case of this comment, both classes are equally true, so it was decided to place the comment on both classes. ...
Article
Full-text available
Purpose User-generated social media comments can be a useful source of information for understanding online corporate reputation. However, the manual classification of these comments is challenging due to their high volume and unstructured nature. The purpose of this paper is to develop a classification framework and machine learning model to overcome these limitations. Design/methodology/approach The authors create a multi-dimensional classification framework for the online corporate reputation that includes six main dimensions synthesized from prior literature: quality, reliability, responsibility, successfulness, pleasantness and innovativeness. To evaluate the classification framework’s performance on real data, the authors retrieve 19,991 social media comments about two Finnish banks and use a convolutional neural network (CNN) to classify automatically the comments based on manually annotated training data. Findings After parameter optimization, the neural network achieves an accuracy between 52.7 and 65.2 percent on real-world data, which is reasonable given the high number of classes. The findings also indicate that prior work has not captured all the facets of online corporate reputation. Practical implications For practical purposes, the authors provide a comprehensive classification framework for online corporate reputation, which companies and organizations operating in various domains can use. Moreover, the authors demonstrate that using a limited amount of training data can yield a satisfactory multiclass classifier when using CNN. Originality/value This is the first attempt at automatically classifying online corporate reputation using an online-specific classification framework.
... Whilst no agreed sample size has been delimited in Twitter sentiment analysis research to date, the final sample fitted with previously set benchmarks (e.g. Canhoto and Padmanabhan, 2015;Lucas et al., 2017;Taecharungroj, 2017;Yu and Wang, 2015) and provided ample data for analysis. ...
... Extensive research has been conducted into the reliability of automated sentiment analysis tools (e.g. Canhoto and Padmanabhan, 2015;Giachanou and Crestani, 2016;Roberts et al., 2018), yielding an emerging consensus that machine-learning and automated sentiment mining and analysis pose a number of risks for researchers. Across a number of automated sentiment analysis tools, significant, consistent errors have been found (Canhoto and Padmanabhan, 2015). ...
... Canhoto and Padmanabhan, 2015;Giachanou and Crestani, 2016;Roberts et al., 2018), yielding an emerging consensus that machine-learning and automated sentiment mining and analysis pose a number of risks for researchers. Across a number of automated sentiment analysis tools, significant, consistent errors have been found (Canhoto and Padmanabhan, 2015). Findings suggest that negative and neutral tweets are more likely to be correctly identified and annotated than positive tweets by automated systems, thus potentially biasing results (Lucas et al., 2017). ...
Article
Purpose The purpose of this paper is to explore consumer attitudes towards ambush marketing and official event sponsorship through the lens of sentiment analysis, and to examine social media users' ethical responses to digital event marketing campaigns during the 2018 FIFA World Cup. Design/methodology/approach The study employed a sentiment analysis, examining Twitter users’ utilization of sponsor and non-sponsor promotional hashtags. Statistical modelling programme R was used to access Twitter’s API, enabling the analysis and coding of user tweets pertaining to six marketing campaigns. The valence of each tweet – as well as the apparent user motivation underlying each post – was assessed, providing insight into Twitter users’ ethical impressions of sponsor and ambush marketer activities on social media and online engagement with social media marketing. Findings The study’s findings indicate that consumer attitudes towards ambush marketing may be significantly more positive than previously thought. Users’ attitudes towards ambush marketing appear significantly more positive than previously assumed, as users of social media emerged as highly responsive to creative and value-added non-sponsor campaigns. Originality/value The findings affirm that sentiment analysis may afford scholars and practitioners a viable means of assessing consumer attitudes towards social marketing activations, dependent upon campaign objectives and strategy. The study provides a new and invaluable context to consumer affect and ambush ethics research, advancing sponsorship and ambush marketing delivery and social sponsorship analytical practice.
... Recent research comparing automated with manual analysis of social media conversations (Canhoto & Padmanabhan, 2015) suggests that automated analysis can be unreliable, especially when sentiment is negative, subtle text elements are used (e.g., emoji, pictures), when meaning is implied through the context rather than made explicit, and when social media users employ irony and sarcasm in their tweets. Our preliminary analysis of the exploratory dataset suggested that these conditions were met in the Protein World firestorm, requiring us to engage in a manual analysis, in line with the qualitative analysis of netnographic data, which pays special attention to rich detail and context. ...
Article
Full-text available
Leading research on social media firestorms typically advises managers to quickly quell the backlash by appeasing brand critics. Drawing on crisis communications and branding research, we offer a radically different perspective and argue that brands can benefit from fighting back online. Through a netnography of a moral-based firestorm, we contribute to the marketing and crisis communications literatures by: identifying the escalation strategy as a way to build brand value; explaining how brands can activate supporters; and providing guidance on how to assess these morally-steeped events. We advance branding research by: identifying how managers can provoke consumer-generated brand stories; and uncovering the hidden benefits of negative consumer voices. Finally, we outline a new perspective on how brands are dialogically constructed through a process we call 'flyting'.
... In addition, the Sentiment Analysis in unstructured texts, typical of social media, is a challenging task, due to the informal nature of these texts, which commonly include abbreviations, spelling errors, emoticons, emojis and informal syntax, that current methods of Sentiment Analysis do not adequately support [9]. ...
... An example of this can be found in Canhoto and Padmanabhan [9] who undertook a comparative study of automated versus manual analysis of social media content. Their results show low levels of agreement between manual and automated classifications, and it proves that, regardless of the automatic method and the advances, there is still much to be done in this area. ...