Search for: All records

Dialog systems (e.g., chatbots) have been widely studied, yet related research that leverages artificial intelligence (AI) and natural language processing (NLP) is constantly evolving. These systems have typically been developed to interact with humans in the form of speech, visual, or text conversation. As humans continue to adopt dialog systems for various objectives, there is a need to involve humans in every facet of the dialog development life cycle for synergistic augmentation of both the humans and the dialog system actors in real-world settings. We provide a holistic literature survey on the recent advancements inhuman-centered dialog systems(HCDS). Specifically, we provide background context surrounding the recent advancements in machine learning-based dialog systems and human-centered AI. We then bridge the gap between the two AI sub-fields and organize the research works on HCDS under three major categories (i.e., Human-Chatbot Collaboration, Human-Chatbot Alignment, Human-Centered Chatbot Design & Governance). In addition, we discuss the applicability and accessibility of the HCDS implementations through benchmark datasets, application scenarios, and downstream NLP tasks. 

Abstract Capsicum chinense (habanero pepper) exhibits substantial variation in fruit pungency, color, and flavor due to its rich secondary metabolite composition, including capsaicinoids, carotenoids, and volatile organic compounds (VOCs). To dissect the genetic and regulatory basis of these traits, we conducted an integrative analysis across 244 diverse accessions using metabolite profiling, genome-wide association studies (GWAS), and transcriptome-wide association studies (TWAS). GWAS identified 507 SNPs for capsaicinoids, 304 for carotenoids, and 1176 for VOCs, while TWAS linked gene expression to metabolite levels, highlighting biosynthetic and regulatory genes in phenylpropanoid, fatty acid, and terpenoid pathways. Segmental RNA sequencing across fruit tissues of contrasting accessions revealed 7034 differentially expressed genes, including MYB31, 3-ketoacyl-CoA synthase, phytoene synthase, and ABC transporters. Notably, AP2 transcription factors and Pentatrichopeptide repeat (PPR) emerged as central regulators, co-expressed with carotenoid and VOC biosynthetic genes. High-resolution spatial transcriptomics (Stereo-seq) identified 74 genes with tissue-specific expression that overlap with GWAS and TWAS loci, reinforcing their regulatory relevance. To validate these candidates, we employed CRISPR/Cas9 to knock out AP2 and PPR genes in tomato. Widely targeted metabolomics and carotenoid profiling revealed major metabolic shifts: AP2 mutants accumulated higher levels of β-carotene and lycopene. In contrast, PPR mutants altered xanthophyll ester and apocarotenoid levels, supporting their roles in carotenoid flux and remodeling. This study provides the first integrative GWAS–TWAS–spatial transcriptomics in C. chinense, revealing key regulators of fruit quality traits. These findings lay the groundwork for precision breeding and metabolic engineering to enhance nutritional and sensory attributes in peppers. 

The adoption of single-use plastics for fabricating lab-on-chip devices used in sensors, chemical and biomedical processes is escalating into a major environmental issue. To address the global need for developing long-term sustainable solutions, we present wood microfluidics as an alternative for electrochemical sensing. The lab-on- wood-chip (LOWC) device developed in this study demonstrated (i) versatility in electrochemical applications (electropolymerization and corrosion analysis), (ii) stability under highly acidic (pH 0.5), basic (pH 14.0) and varied temperature (4◦–60 ◦C) conditions (iii) long-term consistency in performance (>12 months), and (iv) potential for on-field nitrate sensing towards environmental monitoring – in a cost-effective, simple and sustainable manner. 

A significant body of research is dedicated to developing language models that can detect various types of online abuse, for example, hate speech, cyberbullying. However, there is a disconnect between platform policies, which often consider the author's intention as a criterion for content moderation, and the current capabilities of detection models, which typically lack efforts to capture intent. This paper examines the role of intent in the moderation of abusive content. Specifically, we review state-of-the-art detection models and benchmark training datasets to assess their ability to capture intent. We propose changes to the design and development of automated detection and moderation systems to improve alignment with ethical and policy conceptualizations of these abuses. 

Geospatial data visualization on a map is an essential tool for modern data exploration tools. However, these tools require users to manually configure the visualization style including color scheme and attribute selection, a process that is both complex and domain-specific. Large Language Models (LLMs) provide an opportunity to intelligently assist in styling based on the underlying data distribution and characteristics. This paper demonstrates LASEK, an LLM-assisted visualization framework that automates attribute selection and styling in large-scale spatio-temporal datasets. The system leverages LLMs to determine which attributes should be highlighted for visual distinction and even suggests how to integrate them in styling options improving interpretability and efficiency. We demonstrate our approach through interactive visualization scenarios, showing how LLM-driven attribute selection enhances clarity, reduces manual effort, and provides data-driven justifications for styling decisions. 

Abstract Zero‐standby power sensors are crucial for enhancing the safety and widespread adoption of hydrogen (H₂) technologies in chemical processes and sustainable energy applications, given the flammability of H₂at low concentrations. Here, we report an event‐driven hydrogen sensing system utilizing palladium (Pd)‐based micromechanical cantilever switches. The detection mechanism relies on strain generation in the Pd layer, which undergoes reversible volume expansion upon hydrogen adsorption. Our experimental and simulation results demonstrate that the bistable micromechanical switch‐based sensor generates a wake‐up signal with activation time depending on hydrogen concentration in the target environment while always remaining active for events without any standby power consumption under normal conditions. The H₂adsorption‐induced subsequent switching of the multi‐cantilever‐based switch configuration on the sensor resulted in the quasi‐quantification of hydrogen concentrations. The reported zero‐standby power sensor's operational lifetime is limited by the frequency of detection events and exposure to concentrations exceeding hydrogen's flammability limit. This work advances the development of high‐density, maintenance‐free sensor networks for large‐scale deployment with Internet of Things devices, enabling unattended continuous monitoring of hydrogen generation, transportation, distribution, and end‐user applications.