Complex Engineering System Institute

Analyzing social media data has become increasingly important for TV channels as it offers valuable insights into audience preferences, engagement, and sentiment. It provides the feedback needed to align content with audience tastes, fostering greater viewer satisfaction and loyalty. In this paper, we propose a media analytics framework for analyzing audience engagement with posts made by TV channels on their social media platforms. We combine traditional statistical modeling, natural language processing tools, and explainable machine learning to predict engagement and derive feature importance. Our results on two platforms (Facebook and Instagram) for a major TV channel in Chile confirm the predictive capabilities of machine learning: Extreme Gradient Boosting achieved the best performance on Facebook, with a Mean Absolute Percentage Error (MAPE) of 25.14%, while the lowest MAPE for Instagram (16.56%) was obtained using Support Vector Regression. Furthermore, explainable machine learning techniques unveil interesting conclusions for decision-making, such as the importance of mentions of TV personalities, which can be automated using Named Entity Recognition.

Eye-tracking has gained considerable attention across multiple research domains. Recently, web-based eye-tracking has become feasible, demonstrating reliable performance in perceptual and cognitive tasks. However, its systematic evaluation in decision-making remains unknown. Here we compare a laboratory-based eye tracker (the EyeLink 1000 Plus) with a webcam-based method (WebGazer) across two discrete-choice experiments. We systematically manipulated display size to approximate common device classes (monitor, laptop, tablet, mobile) and task complexity (simple vs. complex choice matrices). We find that on larger displays and simpler tasks, WebGazer produces gaze patterns and parameter inferences from computational models of behavior comparable to EyeLink. However, reliability diminishes on smaller displays and with more complex choice matrices. These results provide the first systematic evaluation of web-based eye tracking for decision-making research and offer practical guidance regarding its viability for online behavioral studies.

Lung cancer is a leading cause of death in Chile, where late-stage diagnoses and high mortality rates prevail. Here, we describe the development of OncovigIA, a novel digital tool powered by natural language processing that enhances the identification of potential lung cancer cases by surveilling computed tomography (CT) reports in a large public Hospital in Santiago, Chile. We combined natural language processing and large language models with state-of-the-art machine learning techniques and approaches to treat unbalanced data sets and determine the best solution to implement in OncovigIA. Focusing on key sections of the reports and using various machine learning models, including a balanced Random Forest, the tool achieved high performance with 0.90 accuracy and 0.84 F1-score on the test set. When applied to 13,326 CT chest reports from 2022, it successfully identified 377 CTs of patients with suspected lung cancer previously undetected and not managed by the multidisciplinary local lung cancer team. This study underscores the potential of artificial intelligence in early cancer detection and highlights the importance of its integration into local health care ecosystems. By promptly increasing the number of patients referred for specialized management, the tool OncovigIA offers a promising path toward improving lung cancer survival rates in Chile and beyond. Moreover, this article provides avenues for its broader implementation, extending it to other cancer types and/or health care-related texts for continuous surveillance, aiming at the early referral and treatment of cancer in low-resource settings.

ABSTRACTThis article analyzes the provision of public goods in a two‐player game setting, employing the Game with Observable Delay (GOD) framework to investigate how endogenous timing influences contribution strategies and welfare outcomes. Our analysis shows that, for both symmetric and asymmetric valuation cases, the endogenous timing outcome leads to simultaneous play. This result arises from players' strong aversion to the follower position, driven by the leader's free‐riding incentives. The outcome is good for welfare, since in both the symmetric and asymmetric cases, simultaneous play leads to greater total contributions compared to the sequential equilibrium. If the framework for contributions was of the Game with Action Commitment type, the outcome would be simultaneous, leading to an inferior welfare result. These findings indicate that fostering environments with a structure akin to the GOD could strategically encourage players to maximize their contributions and improve welfare outcomes in public goods scenarios.