Export
Alex Morales
2 previous affiliations
- PhD student · 2014–2020
- MS student · 2012–2014
- Undergrad student · 2008–2012
Research Overview
Alex Morales focuses on natural language processing, specifically analyzing user-generated content for deception, humor, and misinformation. Early research explored the generation and detection of synthetic review spam, while subsequent work investigated computational approaches to identifying humor in online reviews. Most recently, Morales has contributed to the development of datasets for assessing the health risks of medical misinformation on social media.
Research Trajectory
The research trajectory of Alex Morales spans from 2013 to 2020, with a consistent focus on the computational analysis of user-generated text. Early work primarily addressed the integrity and stylistic nuances of online reviews, later expanding into the broader domain of social media misinformation.
In 2013, Morales led research on the vulnerabilities of online review platforms to automated manipulation. As the first author of Synthetic review spamming and defense, Morales investigated the mechanics of low-cost, machine-generated deceptive reviews. This work highlighted the financial incentives behind opinion spam and demonstrated how a simple review synthesis model could use truthful reviews as templates to generate fake content. By exposing the mechanics of automated review synthesis, the research underscored a significant risk to platforms that rely heavily on user feedback for consumer trust.
Continuing the analysis of online reviews, Morales shifted focus toward the pragmatic and stylistic dimensions of user-generated text. In the 2017 paper Identifying Humor in Reviews using Background Text Sources, Morales served as first author, exploring computational methods to detect humor in review text by leveraging external background sources. This research contributed to the broader natural language processing challenge of understanding non-literal and subjective language in digital environments.
By 2020, Morales's collaborative efforts expanded into the critical area of public health misinformation. Participating as a middle author, Morales contributed to Drink bleach or do what now? Covid-HeRA: A dataset for risk-informed health decision making in the presence of COVID19 misinformation. This work addressed the rapid spread of inaccurate medical advice during the COVID-19 pandemic. The research team introduced the Covid-HeRA dataset to facilitate the health risk assessment of social media posts, focusing on evaluating the severity and potential harm of specific misinformation stories and identifying signals that characterize highly malicious fake news.
Current Directions
Based on the most recent paper in the evidence from 2020, Morales's research interests encompass the detection and risk assessment of misinformation in digital spaces. This direction addresses the urgent need for natural language processing tools capable of evaluating the severity of false claims, particularly in high-stakes domains such as public health. The focus lies not only in identifying fake news but also in understanding the potential harm of refuted claims and developing robust datasets to train risk-informed decision-making models.
Profile updated
Loading new papers…
Loading research neighbors…
See something wrong? Report a mistake