Unsupervised Latent Dirichlet Allocation Research Articles

Look What We Have Here: Exploring Brand-Related Sport Consumer Twitter Conversation Topics

Abstract: As sport organizations leverage social media as a critical component of marketing strategy, tools for exploring the large volume of sport consumer social media conversations are vital. This scholarship demonstrates the value of unsupervised latent Dirichlet allocation (LDA) as a tool for exploring consumers' digital conversations. Specifically, unsupervised LDA was applied to derive latent topics among Women's National Basketball Association-related Twitter conversation over the course of the 2020 season. Quantitative (cv and umass scores) and qualitative (two expert reviews) approaches were utilized to delineate topic configurations. Marginal topic distance established topic importance. Results from 118,518 tweets revealed 18 conversation topics spanning two overarching themes: social justice issues and on-court performance. The range and depth of the results highlight the importance of the unsupervised topic modeling method (without semi-supervised predetermined topic leads) for considering holistic rather than subsampled or snapshot datasets. This empirical investigation extends the conversation surrounding natural language processing to sport management research and practice, delivers a foundation for unsupervised LDA application to sport consumer conversation, and explores social media conversations during a critical moment for the WNBA.

A computational approach to cryptocurrency marketing on social media

This study aims to explore social media content associated with cryptocurrency marketing. We employ unsupervised Latent Dirichlet allocation topic modeling and sentiment analysis techniques to 98,716 tweets to examine Twitter (now known as X) content for subjects and sentiments related to cryptocurrency. Our findings reveal that cryptocurrency tweets fell into four categories, with ‘cryptocurrency trading,’ ‘NFT airdrop,’ ‘cryptocurrency affiliate program,’ and ‘Dogecoin on social media’ being the most popular. Furthermore, most of these topics exhibited positive sentiments. This study contributes theoretically by integrating cryptocurrency marketing into the diffusion of innovation paradigm. In addition, it offers strategic insights for digital marketers in identifying prevalent topics and sentiments related to cryptocurrency, enabling the tailoring of affiliate marketing communication strategies on social media.

“Those blimmin Ts and Cs”: a mixed methods analysis of how people manage personal information, privacy, and impressions

ABSTRACT Interconnected and smart technologies complicate personal information management (PIM) because users delegate the storing, organizing, and retrieving of personal information to smart and mobile service providers. Meta-level PIM activities are required to maintain the privacy and security of personal information. This study provides insights into how users of location tracking, mobile apps, and smart home technologies perceive PIM and privacy. We turn to the privacy as contextual integrity (CI) and impression management (IM) literatures to explore informational norms and interpersonal dynamics in PIM. This study is based on a mixed methods design to analyze focus groups and interviews with 106 British and Dutch respondents. Combining unsupervised Latent Dirichlet allocation (LDA) topic modeling and thematic analysis, we reveal discursive patterns in respondent accounts of technology use and provide an in-depth interpretation of these patterns. Our findings indicate that PIM practices are associated with the perceived appropriateness of information flows, anthropomorphic interpretations of technologies, and interpersonal surveillance. Thus, impressions are managed toward social actors as well as technology providers. We contribute to PIM research with a demonstration of how PIM in mobile, smart, and location-based technology use cannot be separated from contextual factors and strategies to manage impressions of habits and behaviors.

"The Smartest Decision for My Future": Social Media Reveals Challenges and Stress During Post-College Life Transition

The post-college transition is a critical period where individuals experience unique challenges and stress before, during, and after graduation. Individuals often use social media to discuss and share information, advice, and support related to post-college challenges in online communities. These communities are important as they fill gaps in institutional support between college and post-college plans. We empirically study the challenges and stress expressed on social media around this transition as students graduate college and move into emerging adulthood. We assembled a dataset of about 299,000 Reddit posts between 2008 and 2020 about the post-college transition from 10 subreddits. We extracted top concerns, challenges, and conversation points using unsupervised Latent Dirichlet Allocation (LDA). Then, we combined the results of LDA with binary transfer learning to identify stress expressions in the dataset (classifier performance at F1=0.94). Finally, we explore temporal patterns in stress expressions, and the variance of per-topic stress levels throughout the year. Our work highlights more deliberate and focused understanding of the post-college transition, as well as useful research and design impacts to study transient cohorts in need of support.

Customer Experience: Extracting Topics From Tweets

The ubiquity of social media platforms facilitates free flow of online chatter related to customer experience. Twitter is a prominent social media platform for sharing experiences, and e-retail firms are rapidly emerging as the preferred shopping destination. This study explores customers’ online shopping experience tweets. Customers tweet about their online shopping experience based on moments of truth shaped by encounters across different touchpoints. We aggregate 25,173 such tweets related to six e-retailers tweeted over a 5-year period. Grounded on agency theory, we extract the topics underlying these customer experience tweets using unsupervised latent Dirichlet allocation. The output reveals five topics which manifest into customer experience tweets related to online shopping—ordering, customer service interaction, entertainment, service outcome failure, and service process failure. Topics extracted are validated through inter-rater agreement with human experts. The study, thus, derives topics from tweets about e-retail customer experience and thereby facilitates prioritization of decision-making pertaining to critical service encounter touchpoints.

A text-based monitoring tool for the legitimacy and guidance of technological innovation systems

Disruptive technologies have to overcome their liability of newness and transition into their growth phase by achieving compliance with existing institutions and pursuing the most promising development paths. Technological innovation system (TIS) studies examined these two issues of legitimacy and guidance of innovation activities by investigating the public discourse with manual media analyses. However, these approaches are time-consuming, prone to subjectivity biases and limited in scope. Therefore, our paper proposes an automatic text analysis methodology based on unsupervised Latent Dirichlet Allocation (LDA) topic modelling and lexicon-based sentiment analysis. By processing 3423 German newspaper articles from the Nexis Uni database, we cover the development of battery-electric vehicles (BEV) in Germany from 2009 to 2019 and identify five socio-technical aspects. Our results indicate an intact legitimacy for the TIS, with Usability, R&D, and Industry being legitimate aspects, which also exhibit strong or improving guidance. In contrast, the Infrastructure and Policy aspects have been less legitimate and weak in guidance, suggesting the need for more holistic policy measures and infrastructure expansion to establish a mass market. Our proposed methodology adds to the toolbox of methods to analyze TIS and serves as a monitoring tool to reveal contested aspects and periods in the public discourse.

Non-Intrusive Assessment of COVID-19 Lockdown Follow-Up and Impact Using Credit Card Information: Case Study in Chile.

In this paper, we propose and validate with data extracted from the city of Santiago, capital of Chile, a methodology to assess the actual impact of lockdown measures based on the anonymized and geolocated data from credit card transactions. Using unsupervised Latent Dirichlet Allocation (LDA) semantic topic discovery, we identify temporal patterns in the use of credit cards that allow us to quantitatively assess the changes in the behavior of the people under the lockdown measures because of the COVID-19 pandemic. An unsupervised latent topic analysis uncovers the main patterns of credit card transaction activity that explain the behavior of the inhabitants of Santiago City. The approach is non-intrusive because it does not require the collaboration of people for providing the anonymous data. It does not interfere with the actual behavior of the people in the city; hence, it does not introduce any bias. We identify a strong downturn of the economic activity as measured by credit card transactions (down to 70%), and thus of the economic activity, in city sections (communes) that were subjected to lockdown versus communes without lockdown. This change in behavior is confirmed by independent data from mobile phone connectivity. The reduction of activity emerges before the actual lockdowns were enforced, suggesting that the population was spontaneously implementing the required measures for slowing virus propagation.

SpaceLDA: Topic distributions aggregation from a heterogeneous corpus for space systems

The design of highly complex systems such as spacecraft entails large amounts of documentation. Tracking relevant information, including hundreds of requirements, throughout several design stages is a challenge. In this study, we propose a novel strategy based on Topic Modelling to facilitate the management of spacecraft design requirements. We introduce spaceLDA, a novel domain-specific semi-supervised Latent Dirichlet Allocation (LDA) model enriched with lexical priors and an optimised Weighted Sum (WS). We collect and curate the first large collection of unstructured data related to space systems, combining several sources: Wikipedia pages, books, and feasibility reports provided by the European Space Agency (ESA). We train the spaceLDA model on three subsets of our heterogeneous training corpus. To combine the resulting per-document topic distributions, we enrich our model with an aggregation method based on an optimised WS. We evaluate our model through a case study, a categorisation of spacecraft design requirements. We finally compare our model’s performance with an unsupervised LDA model and with a literature aggregation method. The results demonstrate that the spaceLDA model successfully identifies the topics of requirements and that our proposed approach surpasses the use of a classic LDA model and the state of the art aggregation method.

COVID-19: Detecting Government Pandemic Measures and Public Concerns from Twitter Arabic Data Using Distributed Machine Learning.

Today’s societies are connected to a level that has never been seen before. The COVID-19 pandemic has exposed the vulnerabilities of such an unprecedently connected world. As of 19 November 2020, over 56 million people have been infected with nearly 1.35 million deaths, and the numbers are growing. The state-of-the-art social media analytics for COVID-19-related studies to understand the various phenomena happening in our environment are limited and require many more studies. This paper proposes a software tool comprising a collection of unsupervised Latent Dirichlet Allocation (LDA) machine learning and other methods for the analysis of Twitter data in Arabic with the aim to detect government pandemic measures and public concerns during the COVID-19 pandemic. The tool is described in detail, including its architecture, five software components, and algorithms. Using the tool, we collect a dataset comprising 14 million tweets from the Kingdom of Saudi Arabia (KSA) for the period 1 February 2020 to 1 June 2020. We detect 15 government pandemic measures and public concerns and six macro-concerns (economic sustainability, social sustainability, etc.), and formulate their information-structural, temporal, and spatio-temporal relationships. For example, we are able to detect the timewise progression of events from the public discussions on COVID-19 cases in mid-March to the first curfew on 22 March, financial loan incentives on 22 March, the increased quarantine discussions during March–April, the discussions on the reduced mobility levels from 24 March onwards, the blood donation shortfall late March onwards, the government’s 9 billion SAR (Saudi Riyal) salary incentives on 3 April, lifting the ban on five daily prayers in mosques on 26 May, and finally the return to normal government measures on 29 May 2020. These findings show the effectiveness of the Twitter media in detecting important events, government measures, public concerns, and other information in both time and space with no earlier knowledge about them.

Target Discrimination Based on Weakly Supervised Learning for High-Resolution SAR Images in Complex Scenes

To design a highly automatic and practical discrimination method for high-resolution synthetic aperture radar (SAR) images in complex scenes, a novel target discrimination framework based on weakly supervised learning (WSL) of the mid-level features is proposed in this article. First, we extract the dense SAR scale-invariant feature transform (SAR-SIFT) features of the candidate regions obtained from the detected SAR images. Then, the dense SAR-SIFT descriptors are transformed into richer mid-level features by coding and pooling. Finally, the mid-level features are input into a WSL-based target discrimination method, where the training set is initially selected by the unsupervised latent Dirichlet allocation (LDA) and iteratively updated by the linear support vector machine (SVM) discriminator. In the proposed method, only the image-level annotations (weak labels), which indicate whether the images containing the targets of interest or not, are required. By introducing WSL, the manual annotations of target regions from SAR images can be avoided, which is generally expensive in complex scenes and may tend to be less accurate and unreliable for the occluded or camouflaged targets. The comprehensive and specific experiments on the measured SAR data have demonstrated the effectiveness of the proposed method in benchmarking with the supervised learning-based linear SVM and linear support vector data description (SVDD) discriminators.

Unsupervised Latent Dirichlet Allocation for supervised question classification

Question answering systems assist users in satisfying their information needs more precisely by providing focused responses to their questions. Among the various systems developed for such a purpose, community-based question answering has recently received researchers’ attention due to the large amount of user-generated questions and answers in social question-and-answer platforms. Reusing such data sources requires an accurate information retrieval component enhanced by a question classifier. The question classification gives the system the possibility to have information about question categories to focus on questions and answers from relevant categories to the input question. In this paper, we propose a new method based on unsupervised Latent Dirichlet Allocation for classifying questions in community-based question answering. Our method first uses unsupervised topic modeling to extract topics from a large amount of unlabeled data. The learned topics are then used in the training phase to find their association with the available category labels in the training data. The category mixture of topics is finally used to predict the label of unseen data.

Public Transit Customer Satisfaction Dimensions Discovery from Online Reviews

Online user-generated content provides a valuable source for identifying dimensions of services. This study proposes a framework for extracting the dimensions of consumer satisfaction of public transportation services using unsupervised latent Dirichlet allocation model. A pilot study was performed on 17,747 online user reviews collected from 1452 public transportation agencies (including streetcar, light rail, heavy rail, boat, and aerial tram) in the United States over 8 years. The proposed approach is able to identify a few dimensions that were not discussed in the previous literature. This research also provides an alternative method to collectively gather users’ feedback and efficiently pre-process textual data related to transit customer satisfaction.

Mining Marketing Meaning from Online Chatter: Strategic Brand Analysis of Big Data Using Latent Dirichlet Allocation

Online chatter, or user-generated content, constitutes an excellent emerging source for marketers to mine meaning at a high temporal frequency. This article posits that this meaning consists of extracting the key latent dimensions of consumer satisfaction with quality and ascertaining the valence, labels, validity, importance, dynamics, and heterogeneity of those dimensions. The authors propose a unified framework for this purpose using unsupervised latent Dirichlet allocation. The sample of user-generated content consists of rich data on product reviews across 15 firms in five markets over four years. The results suggest that a few dimensions with good face validity and external validity are enough to capture quality. Dynamic analysis enables marketers to track dimensions’ importance over time and allows for dynamic mapping of competitive brand positions on those dimensions over time. For vertically differentiated markets (e.g., mobile phones, computers), objective dimensions dominate and are similar across markets, heterogeneity is low across dimensions, and stability is high over time. For horizontally differentiated markets (e.g., shoes, toys), subjective dimensions dominate but vary across markets, heterogeneity is high across dimensions, and stability is low over time.

Extracting Dimensions of Consumer Satisfaction with Quality from Online Chatter: Strategic Brand Analysis of Big Data Using Latent Dirichlet Allocation

Online chatter or User-Generated Content (UGC) constitutes an excellent emerging source for identifying dimensions of quality at a very high temporal frequency. This study proposes a unified framework for extracting the latent dimensions of consumer satisfaction with quality and ascertaining the valence, labels, validity, importance, dynamics, and heterogeneity of those dimensions using unsupervised Latent Dirichlet Allocation (LDA). Our sample of UGC consists of rich data on product reviews across fifteen firms in five markets, over four years. The results suggest that a few dimensions are enough to capture quality. These dimensions have good face and external validity. Dynamic analysis enables tracking the importance of dimensions over time. It also allows for dynamic mapping of competitive brand positions on those dimensions over time. For vertically differentiated markets such as cell phones and computers, objective dimensions dominate and are similar across markets, heterogeneity is low across dimensions, and stability is high across time. For horizontally differentiated markets, such as shoes and toys, subjective dimensions dominate but vary across markets, heterogeneity is high across dimensions, while stability is low over time.