Search results

This study addresses the gap in understanding text analytics within the service domain, focusing on new service development to provide insights into key research themes and trends in text analytics approaches to service development. It explores the benefits and challenges of implementing these approaches and identifies potential research opportunities for future service development. Importantly, this study offers insights to assist service providers to make data-driven decisions for developing new services and optimising existing ones.

This research introduces the hybrid thematic analysis with a systematic literature review (SLR-TA). It delves into the various aspects of text analytics in service development by analysing 124 research papers published from 2012 to 2023. This approach not only identifies key practical applications but also evaluates the benefits and difficulties of applying text analytics in this domain, thereby ensuring the reliability and validity of the findings.

The study highlights an increasing focus on text analytics within the service industry over the examined period. Using the SLR-TA approach, it identifies eight themes in previous studies and finds that “Service Quality” had the most research interest, comprising 42% of studies, while there was less emphasis on designing new services. The study categorises research into four types: Case, Concept, Tools and Implementation, with case studies comprising 68% of the total.

This study is groundbreaking in conducting a thorough and systematic analysis of a broad collection of articles. It provides a comprehensive view of text analytics approaches in the service sector, particularly in developing new services and service innovation. This study lays out distinct guidelines for future research and offers valuable insights to foster research recommendations.

This study aims to introduce an innovative approach to predictive maintenance by integrating time-series sensor data with event logs, leveraging the synergistic potential of deep learning models. The primary goal is to enhance the accuracy of equipment failure predictions, thereby minimizing operational downtime.

The methodology uses a dual-model architecture, combining the patch time series transformer (PatchTST) model for analyzing time-series sensor data and bidirectional encoder representations from transformers for processing textual event log data. Two distinct fusion strategies, namely, early and late fusion, are explored to integrate these data sources effectively. The early fusion approach merges data at the initial stages of processing, while late fusion combines model outputs toward the end. This research conducts thorough experiments using real-world data from wind turbines to validate the approach.

The results demonstrate a significant improvement in fault prediction accuracy, with early fusion strategies outperforming traditional methods by 2.6% to 16.9%. Late fusion strategies, while more stable, underscore the benefit of integrating diverse data types for predictive maintenance. The study provides empirical evidence of the superiority of the fusion-based methodology over singular data source approaches.

This research is distinguished by its novel fusion-based approach to predictive maintenance, marking a departure from conventional single-source data analysis methods. By incorporating both time-series sensor data and textual event logs, the study unveils a comprehensive and effective strategy for fault prediction, paving the way for future advancements in the field.

The Internet has changed consumer decision-making and influenced business behaviour. User-generated product information is abundant and readily available. This paper argues that user-generated content can be efficiently utilised for business intelligence using data science and develops an approach to demonstrate the methods and benefits of the different techniques.

Using Python Selenium, Beautiful Soup and various text mining approaches in R to access, retrieve and analyse user-generated content, we argue that (1) companies can extract information about the product attributes that matter most to consumers and (2) user-generated reviews enable the use of text mining results in combination with other demographic and statistical information (e.g. ratings) as an efficient input for competitive analysis.

The paper shows that combining different types of data (textual and numerical data) and applying and combining different methods can provide organisations with important business information and improve business performance.

The paper shows that combining different types of data (textual and numerical data) and applying and combining different methods can provide organisations with important business information and improve business performance.

The study makes several contributions to the marketing and management literature, mainly by illustrating the methodological advantages of text mining and accompanying statistical analysis, the different types of distilled information and their use in decision-making.

User-generated content was explored to understand the public discourse surrounding the Vessel, a star architecture in New York. Through text analytics, the study aims to uncover topics, sentiments and themes in public opinion regarding this controversial building from social media data.

This study utilized a big data and text analytics approach, employing topic modeling with the BERTopic technique, sentiment analysis with roBERTa and thematic analysis on 10,259 Reddit comments pertaining to the Vessel.

The comments were grouped into 20 topics and seven themes, shedding light on discussions regarding the Vessel’s philosophy of existence, critiques of the architect’s approach, evaluations of project success or failure and considerations of the project’s future. Negative sentiments dominate the discourse, reflecting widespread criticism and skepticism towards the project.

The manual data collection method, due to API restrictions, precluded tracking evolving trends over time. Nevertheless, the study provides insights for architects, urban planners, policymakers and stakeholders involved in public space design and management, highlighting the importance of considering user feedback from social media platforms.

This study enriches our comprehension of how users perceive star architecture in the age of social media, focusing on hidden layers of discourse surrounding a controversial iconic building. By combining topic modeling and sentiment analysis, the study offers a novel approach to analyzing architectural public debates on social media platforms like Reddit.

Technology serves as a key catalyst in shaping society and the economy, significantly altering customer dynamics. Through a deep understanding of these evolving behaviors, a service can be tailored to address each customer's unique needs and personality. We introduce a strategy to integrate customer complaints with their personality traits, enabling responses that resonate with the customer’s unique personality.

We propose a strategy to incorporate customer complaints with their personality traits, enabling responses that reflect the customer’s unique personality. Our approach is twofold: firstly, we employ the customer complaints ontology (CCOntology) framework enforced with multi-class classification based on a machine learning algorithm, to classify complaints. Secondly, we leverage the personality measurement platform (PMP), powered by the big five personality model to predict customer’s personalities. We develop the framework for the Indonesian language by extracting tweets containing customer complaints directed towards Indonesia's three biggest e-commerce services.

By mapping customer complaints and their personality type, we can identify specific personality traits associated with customer dissatisfaction. Thus, personalizing how we offer the solution based on specific characteristics.

The research enriches the state-of-the-art personalizing service research based on captured customer behavior. Thus, our research fills the research gap in considering customer personalities. We provide comprehensive insights by aligning customer feedback with corresponding personality traits extracted from social media data. The result is a highly customized response mechanism attuned to individual customer preferences and requirements.

This paper presents a 10-year systematic review of research on the visual framing of crises to identify the priorities, theories applied and trends in the scholarship of visual framing during crises. The gaps are analyzed to provide evidence-based recommendations for advancing future research.

A total of 269 articles published in 156 peer-reviewed communication journals between January 2014 to December 2023 were reviewed. Data were analyzed using open and axial qualitative coding. A codebook was developed for the quantitative coding and data were analyzed in SPSS descriptive statistics and chi-square tests to answer the research questions.

The proportion of visual framing of crises has remained the same in the last 10 years – there is significantly more research on the visual framing of non-crises. Overall, research on the visual framing of crises is largely exploratory/descriptive and could benefit from a research agenda that is more theory driven. Additionally, there is a skewed focus for research on North America compared to other regions, and for political communication and climate compared to other themes. Environmental sciences and engineering are the most widely investigated journal fields, while disaster is the most common typology studied when looking at the visual framing of crises.

The systematic literature review has some limitations – most particularly that the sample was drawn from a single publisher, which may not be exhaustive enough to represent the full population of articles in the field of visual communication. However, it is a systematic review of the publications that are officially aligned with three of the major communication organizations – the International Communication Association, National Communication Association and World Communication Association. However, future research considering the inclusion of an additional publishers, like Emerald, would further enrich scholarship in visual framing during crises. Second, manual coding of the articles could present potential differences in analysis and interpretation by other researchers. Despite the limitations, the study also provides some important insights into the present and future of the visual framing of crises.

Addressing gaps in the internationalization of visual crisis communication would expand studies for visual framing among underrepresented communities such as populations with low reading literacy, gender minorities and displaced communities and inform visual framing strategies for government and relevant institutions as primary information disseminators during crises.

Addressing the gaps identified in this systematic literature review on the visual framing of crises is important for extending theory in this relatively nascent field and guiding crisis visual framing strategies to mitigate uncertainty and panic, threats to stakeholder relationships, social vulnerabilities and the visual framing of stakeholder-centric crisis responses.

Based on available literature, this is the first systematic literature review investigating the use of all types of visuals used during all crisis typologies, reflecting the ubiquity of crises and the increased focus on the use of visuals in crisis communication in the last decade.

Product returns are a major challenge for e-businesses as they involve huge logistical and operational costs. Therefore, it becomes crucial to predict returns in advance. This study aims to evaluate different genres of classifiers for product return chance prediction, and further optimizes the best performing model.

An e-commerce data set having categorical type attributes has been used for this study. Feature selection based on chi-square provides a selective features-set which is used as inputs for model building. Predictive models are attempted using individual classifiers, ensemble models and deep neural networks. For performance evaluation, 75:25 train/test split and 10-fold cross-validation strategies are used. To improve the predictability of the best performing classifier, hyperparameter tuning is performed using different optimization methods such as, random search, grid search, Bayesian approach and evolutionary models (genetic algorithm, differential evolution and particle swarm optimization).

A comparison of F1-scores revealed that the Bayesian approach outperformed all other optimization approaches in terms of accuracy. The predictability of the Bayesian-optimized model is further compared with that of other classifiers using experimental analysis. The Bayesian-optimized XGBoost model possessed superior performance, with accuracies of 77.80% and 70.35% for holdout and 10-fold cross-validation methods, respectively.

Given the anonymized data, the effects of individual attributes on outcomes could not be investigated in detail. The Bayesian-optimized predictive model may be used in decision support systems, enabling real-time prediction of returns and the implementation of preventive measures.

There are very few reported studies on predicting the chance of order return in e-businesses. To the best of the authors’ knowledge, this study is the first to compare different optimization methods and classifiers, demonstrating the superiority of the Bayesian-optimized XGBoost classification model for returns prediction.

To predict the price of battery-grade lithium carbonate accurately and provide proper guidance to investors, a method called MFTBGAM is proposed in this study. This method integrates textual and numerical information using TCN-BiGRU–Attention.

The Word2Vec model is initially employed to process the gathered textual data concerning battery-grade lithium carbonate. Subsequently, a dual-channel text-numerical extraction model, integrating TCN and BiGRU, is constructed to extract textual and numerical features separately. Following this, the attention mechanism is applied to extract fusion features from the textual and numerical data. Finally, the market price prediction results for battery-grade lithium carbonate are calculated and outputted using the fully connected layer.

Experiments in this study are carried out using datasets consisting of news and investor commentary. The findings reveal that the MFTBGAM model exhibits superior performance compared to alternative models, showing its efficacy in precisely forecasting the future market price of battery-grade lithium carbonate.

The dataset analyzed in this study spans from 2020 to 2023, and thus, the forecast results are specifically relevant to this timeframe. Altering the sample data would necessitate repetition of the experimental process, resulting in different outcomes. Furthermore, recognizing that raw data might include noise and irrelevant information, future endeavors will explore efficient data preprocessing techniques to mitigate such issues, thereby enhancing the model’s predictive capabilities in long-term forecasting tasks.

The price prediction model serves as a valuable tool for investors in the battery-grade lithium carbonate industry, facilitating informed investment decisions. By using the results of price prediction, investors can discern opportune moments for investment. Moreover, this study utilizes two distinct types of text information – news and investor comments – as independent sources of textual data input. This approach provides investors with a more precise and comprehensive understanding of market dynamics.

We propose a novel price prediction method based on TCN-BiGRU Attention for “text-numerical” information fusion. We separately use two types of textual information, news and investor comments, for prediction to enhance the model's effectiveness and generalization ability. Additionally, we utilize news datasets including both titles and content to improve the accuracy of battery-grade lithium carbonate market price predictions.

This chapter focuses on how alternative data can change the nature of financial forecasting through improved short-term forecasting techniques and decreased informativeness from longer term sources. Increased use of social media data leads the charge in transforming this transition. Alternative data are data not from standard financial statements or formal reports. This chapter looks at alternative data from new sources (e.g., social media, Internet of Things [IoT], and digital footprints) and alternative data from new collection methods like web scraping for textual analysis, image analysis, and vocal analysis). It first discusses standard data in financial forecasting. Next, this chapter examines alternative data in financial forecasting. Finally, it discusses alternative data used in studying finance more broadly.

Historical newspaper collections provide a wealth of information about the past. Although the digitization of these collections significantly improves their accessibility, a large portion of digitized historical newspaper collections, such as those of KBR, the Royal Library of Belgium, are not yet searchable at article-level. However, recent developments in AI-based research methods, such as document layout analysis, have the potential for further enriching the metadata to improve the searchability of these historical newspaper collections. This paper aims to discuss the aforementioned issue.

In this paper, the authors explore how existing computer vision and machine learning approaches can be used to improve access to digitized historical newspapers. To do this, the authors propose a workflow, using computer vision and machine learning approaches to (1) provide article-level access to digitized historical newspaper collections using document layout analysis, (2) extract specific types of articles (e.g. feuilletons – literary supplements from Le Peuple from 1938), (3) conduct image similarity analysis using (un)supervised classification methods and (4) perform named entity recognition (NER) to link the extracted information to open data.

The results show that the proposed workflow improves the accessibility and searchability of digitized historical newspapers, and also contributes to the building of corpora for digital humanities research. The AI-based methods enable automatic extraction of feuilletons, clustering of similar images and dynamic linking of related articles.

The proposed workflow enables automatic extraction of articles, including detection of a specific type of article, such as a feuilleton or literary supplement. This is particularly valuable for humanities researchers as it improves the searchability of these collections and enables corpora to be built around specific themes. Article-level access to, and improved searchability of, KBR's digitized newspapers are demonstrated through the online tool (https://tw06v072.ugent.be/kbr/).