As smart meter technology advances, the advanced metering infrastructure (AMI) enables real-time monitoring of electricity usage at 15-minute intervals. Since prior studies mainly focused on industrial infrastructures, anomaly detection in residential electricity usage remains underexplored. In this paper, we propose a gated recurrent unit (GRU)-based deep learning model to detect resident changes in residential households using the electricity metering data through AMI. Monthly consumption patterns and energy consumption are predicted using actual electric energy metering data collected from 846 apartment households in Seoul, Korea. To detect the resident changes, combined losses are computed by integrating the consumption pattern and consumption energy prediction errors. The experimental results show an accuracy of 82.55%, specificity of 100%, and recall of 65.11% when allowing a +-1-month error margin. The proposed deep leaning model demonstrates the practical feasibility of detecting resident changes in residential households and is expected to contribute to the development of smart grid services.

Predicting emotions elicited by news headlines can be challenging as the task is largely influenced by the varying nature of people’s interpretations and backgrounds. Previous works have explored classifying discrete emotions directly from news headlines. We provide a different approach to tackling this problem by utilizing people’s explanations of their emotion, written in free-text, on how they feel after reading a news headline. Using the dataset BU-NEmo+ (Gao et al., 2022), we found that for emotion classification, the free-text explanations have a strong correlation with the dominant emotion elicited by the headlines. The free-text explanations also contain more sentimental context than the news headlines alone and can serve as a better input to emotion classification models. Therefore, in this work we explored generating emotion explanations from headlines by training a sequence-to-sequence transformer model and by using pretrained large language model, ChatGPT (GPT-4). We then used the generated emotion explanations for emotion classification. In addition, we also experimented with training the pretrained T5 model for the intermediate task of explanation generation before fine-tuning it for emotion classification. Using McNemar’s significance test, methods that incorporate GPT-generated free-text emotion explanations demonstrated significant improvement (P-value < 0.05) in emotion classification from headlines, compared to methods that only use headlines. This underscores the value of using intermediate free-text explanations for emotion prediction tasks with headlines.

This study explores the affective responses and newsworthiness perceptions of generative AI for visual journalism. While generative AI offers advantages for newsrooms in terms of producing unique images and cutting costs, the potential misuse of AI-generated news images is a cause for concern. For our study, we designed a 3-part news image codebook for affect-labeling news images based on journalism ethics and photography guidelines. We collected 200 news headlines and images retrieved from a variety of U.S. news sources on the topics of gun violence and climate change, generated corresponding news images from DALL-E 2 and asked annotators their emotional responses to the human-selected and AI-generated news images following the codebook. We also examined the impact of modality on emotions by measuring the effects of visual and textual modalities on emotional responses. The findings of this study provide insights into the quality and emotional impact of generative news images produced by humans and AI. Further, results of this work can be useful in developing technical guidelines as well as policy measures for the ethical use of generative AI systems in journalistic production. The codebook, images and annotations are made publicly available to facilitate future research in affective computing, specifically tailored to civic and public-interest journalism.

This paper introduces a multilingual dataset of COVID-19 vaccine misinformation, consisting of annotated tweets from three middle-income countries: Brazil, Indonesia, and Nigeria.The expertly curated dataset includes annotations for 5,952 tweets, assessing their relevance to COVID-19 vaccines, presence of misinformation, and the themes of the misinformation.To address challenges posed by domain specificity, the low-resource setting, and data imbalance, we adopt two approaches for developing COVID-19 vaccine misinformation detection models: domain-specific pre-training and text augmentation using a large language model.Our best misinformation detection models demonstrate improvements ranging from 2.7 to 15.9 percentage points in macro F1-score compared to the baseline models.Additionally, we apply our misinformation detection models in a large-scale study of 19 million unlabeled tweets from the three countries between 2020 and 2022, showcasing the practical application of our dataset and models for detecting and analyzing vaccine misinformation in multiple countries and languages.Our analysis indicates that percentage changes in the number of new COVID-19 cases are positively associated with COVID-19 vaccine misinformation rates in a staggered manner for Brazil and Indonesia, and there are significant positive associations between the misinformation rates across the three countries.

Modern embedded systems are increasingly complex and contain multiple software layers from BSP (Board Support Packages) to OS to middleware to AI (Artificial Intelligence) algorithms like perception and voice recognition. Integrations of inter-layer and intra-layer in embedded systems provide dedicated services such as taking a picture or movie-streaming. Accordingly, it gets more complicated to find out the root cause of a system failure. This industrial proposal describes a difficulty of testing embedded systems, and presents a case study in terms of integration testing.

In a wireless system, the relay station (RS) is introduced to increase the data rate of a mobile station (MS) in a shadowing area or a cell edge. RS is stationary of non-stationary and takes charge of relaying between a base station (BS) and an MS. However, the adoption of several RSs causes inefficient time-slot. In this paper we propose a novel algorithm for utilizing RSs and provide an appropriate frame structure. We divide the zone into two, inner and outer zone for a sectorized cell structure. RSs are stationary at inner zone edge in each sector and connect an MS at outer zone with a BS. And a special time-slot is shared by all RSs existing in the same sector. The proposed algorithm will increase the opportunity of connection between a BS and an MS out of RS coverage. Simulation result shows that throughout increases about three times in inner and outer zone.

In this letter, an analysis on CDMA forward link capacity with hard and soft handoffs, when a handoff decision is based on filtered pilot signal strengths, is provided. For the soft handoff, capacity based on three different power allocation schemes is investigated: equal power, equal signal to interference ratio and selection diversity. Contrary to previous results that soft handoff provides a capacity decrease in the CDMA forward link, it is concluded that soft handoff mitigates capacity loss due to filtering through diversity gain, and may in fact provide higher capacity.

WCDMA handover algorithms employ signal averaging, hysteresis and the time-to-trigger mechanism to optimize the trade off between number of unnecessary handovers, reported events (system load) and handover delay time. We investigate optimal parameters for the WCDMA intra-frequency handover algorithm and the impact of each parameter on the system performance. The number of reporting events triggered for handover and handover delay are key performance measures in this paper. The study shows various tradeoffs between the parameters related to averaging, hysteresis and time-to-trigger. We have also discovered that the layer3 filter and time-to-trigger mechanism may cause negative effects on each other in some cases and there are optimum values, when used simultaneously.