Mario Andrés Paredes-Valverde

Automatic summarization is essential for processing the vast quantity of news articles. However, existing methods struggle with factual consistency, hallucinations, and English-centric evaluations. This paper investigates whether Large Language Models can generate coherent and factually grounded summaries of Spanish-language news articles, using the DACSA dataset as a benchmark. Several strategies are evaluated, including zero-shot prompting, one-shot prompting, fine-tuning of seq2seq models mBART and mT5, and a novel bottleneck prompting method that integrates attention-based salience scoring with Named Entity Recognition. Our results show that modern instruction-tuned language models can achieve competitive performance in zero- and one-shot settings, often approaching the performance of fine-tuned baselines. Our proposed bottleneck method enhances factual accuracy and content selection, leading to measurable improvements in ROUGE and BERTScore, especially for larger models such as LLaMA-3.1-70B and Gemma-2-9B. These results suggest that structured prompting can complement conventional approaches, offering an effective and cost-efficient alternative to full supervision. The results indicate that LLMs guided by entity-anchored bottlenecks provide a promising approach to multilingual summarization in domains with limited resources.

The proliferation of social media has made Sentiment Analysis an essential tool for understanding user opinions, particularly in underrepresented language variants such as Mexican Spanish. Recent advances in Large Language Models have made effective sentiment analysis through in-context learning techniques, reducing the need for supervised training. This study compares the performance of zero and few-shot with traditional fine-tuning approaches of tourism-related texts in Mexican Spanish. Two annotated datasets from the REST-MEX 2022 and 2023 shared tasks were used for this purpose. Results show that fine-tuning, particularly with the MarIA model, achieves the best overall performance. However, modern LLMs that use in-context learning strategies, such as Mixtral 8x7B for zero-shot and Mistral 7B for few-shot, demonstrate strong potential in low-resource settings by closely approximating the accuracy of fine-tuned models, suggesting that in-context learning is a viable alternative to fine-tuning for sentiment analysis in Mexican Spanish when labeled data is limited. These approaches can enable intelligent, data-driven digital services with applications in tourism platforms and urban information systems that enhance user experience and trust in large-scale socio-technical ecosystems.

The early detection of depression in a person is of great help to medical specialists since it allows for better treatment of the condition. Social networks are a promising data source for identifying individuals who are at risk for this mental disease, facilitating timely intervention and thereby improving public health. In this frame of reference, we propose an NLP-based system called Mental-Health for detecting users’ depression levels through comments on X. Mental-Health is supported by a model comprising four stages: data extraction, preprocessing, emotion detection, and depression diagnosis. Using a natural language processing tool, the system correlates emotions detected in users’ posts on X with the symptoms of depression and provides specialists with the depression levels of the patients. By using Mental-Health, we described a case study involving real patients, and the evaluation process was carried out by comparing the results obtained using Mental-Health with those obtained through the application of the PHQ-9 questionnaire. The system identifies moderately severe and moderate depression levels with good precision and recall, allowing us to infer the model’s good performance and confirm that it is a promising option for mental health support.

Internet of Things (IoT) technologies can greatly benefit from machine-learning techniques and artificial neural networks for data mining and vice versa. In the agricultural field, this convergence could result in the development of smart farming systems suitable for use as decision support systems by peasant farmers. This work presents the design of a smart farming system for crop production, which is based on low-cost IoT sensors and popular data storage services and data analytics services on the cloud. Moreover, a new data-mining method exploiting climate data along with crop-production data is proposed for the prediction of production volume from heterogeneous data sources. This method was initially validated using traditional machine-learning techniques and open historical data of the northeast region of the state of Puebla, Mexico, which were collected from data sources from the National Water Commission and the Agri-food Information Service of the Mexican Government.

Among mental health diseases, depression is one of the most severe, as it often leads to suicide; due to this, it is important to identify and summarize existing evidence concerning depression sign detection research on social media using the data provided by users. This review examines aspects of primary studies exploring depression detection from social media submissions (from 2016 to mid-2021). The search for primary studies was conducted in five digital libraries: ACM Digital Library, IEEE Xplore Digital Library, SpringerLink, Science Direct, and PubMed, as well as on the search engine Google Scholar to broaden the results. Extracting and synthesizing the data from each paper was the main activity of this work. Thirty-four primary studies were analyzed and evaluated. Twitter was the most studied social media for depression sign detection. Word embedding was the most prominent linguistic feature extraction method. Support vector machine (SVM) was the most used machine-learning algorithm. Similarly, the most popular computing tool was from Python libraries. Finally, cross-validation (CV) was the most common statistical analysis method used to evaluate the results obtained. Using social media along with computing tools and classification methods contributes to current efforts in public healthcare to detect signs of depression from sources close to patients.

Internet of Things (IoT) technologies can greatly benefit from machine learning techniques and Artificial Neural Networks for data mining and vice versa. In the agricultural field, this convergence could result in the development of smart farming systems suitable for use as decision support systems by peasant farmers. This work presents the design of a smart farming system for crop production, which is based on low-cost IoT sensors and popular data storage services and data analytics services on the Cloud. Moreover, a new data mining method exploiting climate data along with crop production data is proposed for the prediction of production volume from heterogeneous data sources. This method was initially validated using traditional machine learning techniques and open historical data of the northeast region of the state of Puebla, Mexico, which were collected from data sources from the National Water Commission and the Agri-food Information Service of the Mexican Government.