Hasil untuk "Advertising"

Mohammad Mosaffa, Omid Rafieian

Firms collect vast amounts of behavioral and geographical data on individuals. While behavioral data captures an individual's digital footprint, geographical data reflects their physical footprint. Given the significant privacy risks associated with combining these data sources, it is crucial to understand their respective value and whether they act as complements or substitutes in achieving firms' business objectives. In this paper, we combine economic theory, machine learning, and causal inference to quantify the value of geographical data, the extent to which behavioral data can substitute for it, and the mechanisms through which it benefits firms. Using data from a leading in-app advertising platform in a large Asian country, we document that geographical data is most valuable in the early cold-start stage, when behavioral histories are limited. In this stage, geographical data complements behavioral data, improving targeting performance by almost 20%. As users accumulate richer behavioral histories, however, the role of geographical data shifts: it becomes largely substitutable, as behavioral data alone captures the relevant heterogeneity. These results highlight a central privacy-utility trade-off in ad personalization and inform managerial decisions about when location tracking creates value.

Hieu Pham Dinh, Hung Nguyen Huy, Mo El-Haj

VietJobs is the first large-scale, publicly available corpus of Vietnamese job advertisements, comprising 48,092 postings and over 15 million words collected from all 34 provinces and municipalities across Vietnam. The dataset provides extensive linguistic and structured information, including job titles, categories, salaries, skills, and employment conditions, covering 16 occupational domains and multiple employment types (full-time, part-time, and internship). Designed to support research in natural language processing and labour market analytics, VietJobs captures substantial linguistic, regional, and socio-economic diversity. We benchmark several generative large language models (LLMs) on two core tasks: job category classification and salary estimation. Instruction-tuned models such as Qwen2.5-7B-Instruct and Llama-SEA-LION-v3-8B-IT demonstrate notable gains under few-shot and fine-tuned settings, while highlighting challenges in multilingual and Vietnamese-specific modelling for structured labour market prediction. VietJobs establishes a new benchmark for Vietnamese NLP and offers a valuable foundation for future research on recruitment language, socio-economic representation, and AI-driven labour market analysis. All code and resources are available at: https://github.com/VinNLP/VietJobs.

Nesrine MZID

This study investigates the impact of artificial intelligence (AI)-powered advertising on brand equity and purchase intentions among Tunisian webroomers. Drawing on flow theory and the theory of reasoned action, we develop a conceptual model linking AI-based advertising to consumer responses. A survey of 350 Tunisian students was conducted after exposing participants to an AI-powered advertising campaign, and the data were analyzed using structural equation modeling (SEM). The findings confirm that AI-powered advertising significantly enhances both brand equity and webroomers’ purchase intentions. Reliability and validity tests further support the robustness of the measurement model. By addressing an underexplored context in North Africa, this research contributes to the literature on AI in digital marketing and consumer behaviour. The results also provide meaningful managerial implications for companies seeking to leverage AI-based advertising to strengthen brand equity and stimulate purchase intentions.

Yezid Alfonso Cancino Gómez, Pedro Mauricio Torres Duque, Laima Catherine Alfonso Orjuela et al.

Facebook, as a social networking platform, has become a significant marketing tool, serving both as an advertising medium and a sales channel. Considering the diversity in the purchasing behaviors of users, this research aimed to identify patterns that can be used to predict buyer types on Facebook. In order to effectively determine the relevant variables and predictors influencing buyer behavior on this platform, data were collected through a survey administered to 663 participants. Accordingly, an exploratory factor analysis was first conducted, followed by a multimodal logistic regression. The obtained results led to the acceptance of two out of the four proposed hypotheses. Based on the observations made, the majority of the demographic variables were found to be poor predictors of buyer type except for perceived security and responsiveness to advertising. These variables were observed to be strong predictors, significantly influencing the categorization of buyers within the Facebook environment.

Jelena Krstić, Tamara Rajić, Milica Kostić-Stanković et al.

The objective of the study is to examine the influential factors of female consumers’ environmentally conscious apparel purchase behaviour, in the context of an emerging European economy. The study builds upon Stimulus-Organism-Response (S-O-R) theoretical framework and proposes the moderating role of style orientation on the direct effects on purchase behaviour. Online questionnaire-based survey was performed to gather data, using convenience sampling framework. Hypothesized relationships were examined on a sample including 343 responses. Partial Least Squares Structural Equation Modeling (PLS-SEM), using SmartPLS4, was applied to examine proposed research framework. Results of the study point to green self-identity as more influential direct determinant of purchase behaviour, in comparison to green advertising, and the mediating role of green self-identity in the relationship between green advertising and environmentally conscious apparel purchase behaviour. The study revealed the moderating function of style orientation in the relation between green self-identity and purchase behaviour. Theoretical and practical implications of the study are discussed and limitations of the study, followed by future research directions to overcome the drawbacks of the present research, are noted.

Hongyu Yang, Chunxi Wen, Jiyin Zhang et al.

Click-through rates prediction is critical in modern advertising systems, where ranking relevance and user engagement directly impact platform efficiency and business value. In this project, we explore how to model CTR more effectively using a large-scale Taobao dataset released by Alibaba. We start with supervised learning models, including logistic regression and Light-GBM, that are trained on static features such as user demographics, ad attributes, and contextual metadata. These models provide fast, interpretable benchmarks, but have limited capabilities to capture patterns of behavior that drive clicks. To better model user intent, we combined behavioral data from hundreds of millions of interactions over a 22-day period. By extracting and encoding user action sequences, we construct representations of user interests over time. We use deep learning models to fuse behavioral embeddings with static features. Among them, multilayer perceptrons (MLPs) have achieved significant performance improvements. To capture temporal dynamics, we designed a Transformer-based architecture that uses a self-attention mechanism to learn contextual dependencies across behavioral sequences, modeling not only what the user interacts with, but also the timing and frequency of interactions. Transformer improves AUC by 2.81 % over the baseline (LR model), with the largest gains observed for users whose interests are diverse or change over time. In addition to modeling, we propose an A/B testing strategy for real-world evaluation. We also think about the broader implications: personalized ad targeting technology can be applied to public health scenarios to achieve precise delivery of health information or behavior guidance. Our research provides a roadmap for advancing click-through rate predictions and extending their value beyond e-commerce.

Adam Breuer, Bryce J. Dietrich, Michael H. Crespin et al.

This paper introduces the largest and most comprehensive dataset of US presidential campaign television advertisements, available in digital format. The dataset also includes machine-searchable transcripts and high-quality summaries designed to facilitate a variety of academic research. To date, there has been great interest in collecting and analyzing US presidential campaign advertisements, but the need for manual procurement and annotation led many to rely on smaller subsets. We design a large-scale parallelized, AI-based analysis pipeline that automates the laborious process of preparing, transcribing, and summarizing videos. We then apply this methodology to the 9,707 presidential ads from the Julian P. Kanter Political Commercial Archive. We conduct extensive human evaluations to show that these transcripts and summaries match the quality of manually generated alternatives. We illustrate the value of this data by including an application that tracks the genesis and evolution of current focal issue areas over seven decades of presidential elections. Our analysis pipeline and codebase also show how to use LLM-based tools to obtain high-quality summaries for other video datasets.

Anton Fenik, Ernest R. Cadotte, Helena F. Allman

Only a few empirical studies examine the influence of organizational improvisation on firm performance. Further, more insight is needed regarding the coexistence of normative (i.e., planned) and descriptive (i.e., improvised) forms of decision-making. This paper contributes to the literature by filling these gaps. It investigates the moderating effect of entrepreneurial behaviors on the relationship between improvisation and firm performance. It explores whether improvisational decision-making can improve firm performance beyond planning’s effect on firm outcomes. We employ decision theory and sensemaking literature in our conceptual model. We test our hypotheses with data from a cross-sectional managerial survey and a behavioral simulation involving a new product development context. The results reveal a positive moderating effect of entrepreneurial behaviors on the relationship between improvisation and firm performance and that organizational improvisation positively affects firm performance beyond the effect of formal planning.

Amani S. Aljohi, Sara S. Alzaabi, Rahma S. Almahri et al.

This work sheds light on the effectiveness of digital kiosks in targeting specific audiences in contrast to centrally managed mobile phone applications. To this end, we have conducted a case study where a digital kiosk was developed to support the academic activities of the computer science department. Our results show that the students continue to use the mobile phone application. However, the digital kiosk added the following main benefits to the service: Firstly, being in a physical location and thanks to their larger screens, digital kiosks are ‘eye-catching’ devices, which makes them ideal for advertising products/services or communicating relevant information. Secondly, they are brilliant points of attraction. By seeing other people standing in front of any of them, members of the target audience are encouraged to imitate them, even if they did not have the intention to do so. Thirdly, even if the services are available from a mobile phone application, some people do not wish to create an account, download and install the application on their devices, and/or give permission to it, which can potentially invade their privacy and security. Lastly, and equally important, digital kiosks are human-centered technologies that can be more appealing to people who seek social interactions. With this, we conclude that digital kiosks cannot replace mobile phone applications. Rather, they are further technologies that enhance self-service overall.

Abul Ehtesham, Saket Kumar, Aditi Singh et al.

Generative AI is reshaping the media landscape, enabling unprecedented capabilities in video creation, personalization, and scalability. This paper presents a comprehensive SWOT analysis of Metas Movie Gen, a cutting-edge generative AI foundation model designed to produce 1080p HD videos with synchronized audio from simple text prompts. We explore its strengths, including high-resolution video generation, precise editing, and seamless audio integration, which make it a transformative tool across industries such as filmmaking, advertising, and education. However, the analysis also addresses limitations, such as constraints on video length and potential biases in generated content, which pose challenges for broader adoption. In addition, we examine the evolving regulatory and ethical considerations surrounding generative AI, focusing on issues like content authenticity, cultural representation, and responsible use. Through comparative insights with leading models like DALL-E and Google Imagen, this paper highlights Movie Gens unique features, such as video personalization and multimodal synthesis, while identifying opportunities for innovation and areas requiring further research. Our findings provide actionable insights for stakeholders, emphasizing both the opportunities and challenges of deploying generative AI in media production. This work aims to guide future advancements in generative AI, ensuring scalability, quality, and ethical integrity in this rapidly evolving field.

Hamid Eghbalzadeh, Shuai Shao, Saurabh Verma et al.

A myriad of factors affect large scale ads delivery systems and influence both user experience and revenue. One such factor is proactive detection and calibration of seasonal advertisements to help with increasing conversion and user satisfaction. In this paper, we present Proactive Detection and Calibration of Seasonal Advertisements (PDCaSA), a research problem that is of interest for the ads ranking and recommendation community, both in the industrial setting as well as in research. Our paper provides detailed guidelines from various angles of this problem tested in, and motivated by a large-scale industrial ads ranking system. We share our findings including the clear statement of the problem and its motivation rooted in real-world systems, evaluation metrics, and sheds lights to the existing challenges, lessons learned, and best practices of data annotation and machine learning modeling to tackle this problem. Lastly, we present a conclusive solution we took during this research exploration: to detect seasonality, we leveraged Multimodal LLMs (MLMs) which on our in-house benchmark achieved 0.97 top F1 score. Based on our findings, we envision MLMs as a teacher for knowledge distillation, a machine labeler, and a part of the ensembled and tiered seasonality detection system, which can empower ads ranking systems with enriched seasonal information.

Nathalia Beatriz Lobo da Silva, Roseany Patrícia da Silva Rocha, Amanda Cristina de Souza Andrade et al.

ABSTRACT Objective: To analyze the use of ivermectin as COVID-19 prevention method by the population of Mato Grosso in 2020. Methods: This is a home-based survey, carried out between September and October 2020, in 10 pole cities of the socioeconomic regions of State. The use of ivermectin was evaluated through the question: “Did you take ivermectin to prevent COVID-19?”. Sociodemographic variables (sex, age group, education, family income), current work situation, being benefitted by government financial programs, as well as symptoms, seroprevalence of antibodies against SARS-CoV-2, and previous diagnosis of COVID-19 were evaluated. Prevalence and their associations were estimated using the chi-square test. Results: 4.206 individuals were evaluated for prevalence of ivermectin use; 58.3% of the individuals responded positively, this rate being higher in the municipalities of the western region of the state (66.6%). There was no significant difference between sexes, but the prevalence was higher among people aged 50–59 years (69.7%), who were white (66.5%), with complete higher education or more (68.8%) and higher family income (≥3 minimum wages-64.2%). The use of this drug was even higher among participants who considered their knowledge of the disease good or very good (65.0%), who reported having symptoms of COVID-19 (75.3%), and who had been previously diagnosed with the disease (91.2%). Conclusion: There was a high prevalence of use of ivermectin as a method to prevent covid-19 by the population of Mato Grosso, indicating the need for strategies to inform the population about the risks of off-label use of drugs and to combat the advertising of drugs that are ineffective against COVID-19.

Daniel Ladwig, Bianca Lamm, Janis Keuper

In this paper, we describe a first publicly available fine-grained product recognition dataset based on leaflet images. Using advertisement leaflets, collected over several years from different European retailers, we provide a total of 41.6k manually annotated product images in 832 classes. Further, we investigate three different approaches for this fine-grained product classification task, Classification by Image, by Text, as well as by Image and Text. The approach "Classification by Text" uses the text extracted directly from the leaflet product images. We show, that the combination of image and text as input improves the classification of visual difficult to distinguish products. The final model leads to an accuracy of 96.4% with a Top-3 score of 99.2%. We release our code at https://github.com/ladwigd/Leaflet-Product-Classification.

Soraia Teles, Ana Ferreira, Constança Paúl

Abstract Background iSupport is an online program developed by the World Health Organization to provide education, skills training, and social support to informal carers of persons with dementia. This pilot study examines the feasibility of the protocol for a main effectiveness trial of iSupport-Portugal and explores how the intervention and control arms compare over time on well-being outcomes. Methods A mixed-methods experimental parallel between-group design with two arms is followed. Participants were recruited nationwide, by referral or advertising, through the National Alzheimer’s Association. Inclusion criteria are being Portuguese adults, providing e-consent, providing unpaid care to someone with dementia for at least 6 months, experiencing relevant scores on burden (≥ 21 on ZBI) or depression or anxiety (≥ 8 on HADS), and using webpages autonomously. Participants were consecutively randomized to receive iSupport-Portugal or an education-only e-book and were not blinded to group assignment. Data were collected online with self-administered instruments, at baseline, 3 and 6 months after. Outcomes comprise caregiver burden, depression, anxiety, QoL, positive aspects of caregiving, and self-efficacy. Generalized estimating equations were used to estimate group, time, and group-by-time effects. Intervention engagement data were extracted from iSupport’s platform. Semi-structured interviews were conducted. Results Forty-two participants were allocated to the intervention (N = 21) and control (N = 21) arms. Participation (78.1%) and retention rates (73.8%) were fair. More carers in the control arm completed the study (N = 20, 95.2%) than in the intervention arm (N = 11; 52.4%) (χ 2 = 9.98, p = .002). Non-completers were younger, spent less time caring, and scored higher on anxiety. Among carers in the intervention arm, the average attendance rate was of 53.7%. At post-test 38.9% of participants still used iSupport; the remainder participants interrupted use within 2 weeks (Mdn). For per-protocol analyses, significant group-by-time interaction effects favouring the intervention were found for anxiety (Wald χ2 = 6.17, p = .046) and for environmental QoL (Wald χ 2 = 7.06, p = .029). Those effects were not observed in intention-to-treat analyses adjusted for age. Interviewees from the intervention arm (N = 12) reported positive results of iSupport on knowledge and on experiencing positive feelings. No adverse effects were reported. Conclusions This study provides information for a forthcoming full-scale effectiveness trial, as on the acceptability and potential results of iSupport-Portugal. iSupport is suggested as a relevant resource for Portuguese carers. Trial registration ClinicalTrials.gov, NCT04104568 . 26/09/2019.

GP van Rheede van Oudtshoorn, RS ORR

In an over-saturated market, advertisements have become more risqué as companies vie for consumer attention and lesbian content in advertising seems to be on the increase in mainstream media. This article attempts to discover whether lesbian content in advertising elicits positive or negative consumer attitudes towards the advertisement and the brand, and to link these attitudes with the intention to purchase the product. By doing so, marketers will be able to ascertain whether this type of advertising appeal is effective or whether it offends consumers and therefore decreases product sales. The study was quantitative in nature and used descriptive research in a field setting. It was found that there is a significant correlation between tolerance of homosexuality and acceptance of lesbian content in advertising. In addition, these advertisements attracted attention and interest and were not perceived as particularly immoral, exploitive or offensive by most of the sample population. In terms of attracting attention and interest, and being memorable to consumers, advertisements containing clear lesbian interaction are more effective than those with lower levels of homoerotic t

Lucília Cardoso, Meng-Mei Chen, Arthur Araújo et al.

(1) Background: Using neuroscience to understand and influence consumer behavior often leads to ethical controversy. Thus, it is necessary to demystify the use of neuroscience for marketing purposes; the present paper, by accessing the worldwide academic performance in this domain, fulfills this objective. (2) Methods: All extant literature on neuromarketing indexed to the Scopus database—318 articles—was subjected to a bibliometric analysis through a mixed-method approach. (3) Results: The results show that Spain leads the ranks of the most productive countries, while Italian researchers clearly dominate in terms of collaboration. Regarding the most prominent topics, the connection between “Neuroscience” and “Advertising” is highlighted. The findings provide a better understanding of the state-of-the-art in neuromarketing studies, research gaps, and emerging research topics, and additionally provide a new methodological contribution by including SciVal topic prominence in the bibliometric analysis. (4) Conclusions: As practical implications, this study provides useful insights for neuromarketing researchers seeking funding opportunities, which are normally associated with topics within the top prominence percentile or emerging topics. In terms of originality, this study is the first to apply SciVal topic prominence to a bibliometric analysis of neuromarketing, and provides a new bibliometric indicator for neuromarketing research.

Xiao Wang, Shaoguo Liu, Yidong Jia et al.

In the Real-Time Bidding (RTB), advertisers are increasingly relying on bid optimization to gain more conversions (i.e trade or arrival). Currently, the efficiency of bid optimization is still challenged by the (1) sparse feedback, (2) the budget management separated from the optimization, and (3) absence of bidding environment modeling. The conversion feedback is delayed and sparse, yet most methods rely on dense input (impression or click). Furthermore, most approaches are implemented in two stages: optimum formulation and budget management, but the separation always degrades performance. Meanwhile, absence of bidding environment modeling, model-free controllers are commonly utilized, which perform poorly on sparse feedback and lead to control instability. We address these challenges and provide the Multi-Constraints with Merging Features (MCMF) framework. It collects various bidding statuses as merging features to promise performance on the sparse and delayed feedback. A cost function is formulated as dynamic optimum solution with budget management, the optimization and budget management are not separated. According to the cost function, the approximated gradients based on the Hebbian Learning Rule are capable of updating the MCMF, even without modeling of the bidding environment. Our technique performs the best in the open dataset and provides stable budget management even in extreme sparsity. The MCMF is applied in our real RTB production and we get 2.69% more conversions with 2.46% fewer expenditures.

Gagan Aggarwal, Kshipra Bhawalkar, Aranyak Mehta et al.

Internet ad auctions have evolved from a few lines of text to richer informational layouts that include images, sitelinks, videos, etc. Ads in these new formats occupy varying amounts of space, and an advertiser can provide multiple formats, only one of which can be shown. The seller is now faced with a multi-parameter mechanism design problem. Computing an efficient allocation is computationally intractable, and therefore the standard Vickrey-Clarke-Groves (VCG) auction, while truthful and welfare-optimal, is impractical. In this paper, we tackle a fundamental problem in the design of modern ad auctions. We adopt a ``Myersonian'' approach and study allocation rules that are monotone both in the bid and set of rich ads. We show that such rules can be paired with a payment function to give a truthful auction. Our main technical challenge is designing a monotone rule that yields a good approximation to the optimal welfare. Monotonicity doesn't hold for standard algorithms, e.g. the incremental bang-per-buck order, that give good approximations to ``knapsack-like'' problems such as ours. In fact, we show that no deterministic monotone rule can approximate the optimal welfare within a factor better than $2$ (while there is a non-monotone FPTAS). Our main result is a new, simple, greedy and monotone allocation rule that guarantees a $3$ approximation. In ad auctions in practice, monotone allocation rules are often paired with the so-called Generalized Second Price (GSP) payment rule, which charges the minimum threshold price below which the allocation changes. We prove that, even though our monotone allocation rule paired with GSP is not truthful, its Price of Anarchy (PoA) is bounded. Under standard no overbidding assumption, we prove a pure PoA bound of $6$ and a Bayes-Nash PoA bound of $\frac{6}{(1 - \frac{1}{e})}$. Finally, we experimentally test our algorithms on real-world data.

Ziqiao Ao, Gergely Horvath, Chunyuan Sheng et al.

In this paper, we compare different methods to extract skill requirements from job advertisements. We consider three top-down methods that are based on expert-created dictionaries of keywords, and a bottom-up method of unsupervised topic modeling, the Latent Dirichlet Allocation (LDA) model. We measure the skill requirements based on these methods using a U.K. dataset of job advertisements that contains over 1 million entries. We estimate the returns of the identified skills using wage regressions. Finally, we compare the different methods by the wage variation they can explain, assuming that better-identified skills will explain a higher fraction of the wage variation in the labor market. We find that the top-down methods perform worse than the LDA model, as they can explain only about 20% of the wage variation, while the LDA model explains about 45% of it.

Alex Berke, Dan Calacci

In 2020, Google announced it would disable third-party cookies in the Chrome browser to improve user privacy. In order to continue to enable interest-based advertising while mitigating risks of individualized user tracking, Google proposed FLoC. The FLoC algorithm assigns users to "cohorts" that represent groups of users with similar browsing behaviors so that ads can be served to users based on their cohort. In 2022, after testing FLoC in a real world trial, Google canceled the proposal with little explanation. In this work, we provide a post-mortem analysis of two critical privacy risks for FloC by applying an implementation of FLoC to a browsing dataset collected from over 90,000 U.S. devices over a one year period. First, we show how, contrary to its privacy goals, FLoC would have enabled cross-site user tracking by providing a unique identifier for users available across sites, similar to the third-party cookies FLoC was meant to be an improvement over. We show how FLoC cohort ID sequences observed over time can provide this identifier to trackers, even with third-party cookies disabled. We estimate the number of users in our dataset that could be uniquely identified by FLoC IDs is more than 50% after 3 weeks and more than 95% after 4 weeks. We also show how these risks increase when cohort data are combined with browser fingerprinting, and how our results underestimate the true risks FLoC would have posed in a real-world deployment. Second, we examine the risk of FLoC leaking sensitive demographic information. Although we find statistically significant differences in browsing behaviors between demographic groups, we do not find that FLoC significantly risks exposing race or income information about users in our dataset. Our contributions provide insights and example analyses for future approaches that seek to protect user privacy while monetizing the web.