The use of Digital Twins in the industry has become a growing trend recent years, allowing to improve lifecycle any process by taking advantage relationship between physical and virtual world. Existing literature formulates several challenges for building Twins, as well some proposals overcoming them. However, vast majority cases, architectures technologies presented are strongly bounded domain where applied. This article proposes FIWARE Ecosystem, combining its catalog components Smart Data...

This column advocates for including artificial intelligence (AI)-specific metadata on those academic papers that are written with the help of AI in an attempt to analyze use such tools disseminating research.

One of the most widely used methods to evaluate LLMs are Multiple Choice Question (MCQ) tests. MCQ benchmarks enable testing LLM knowledge on almost any topic at scale as results can be processed automatically. To help answer, a few examples called shots included in prompt. Moreover, asked answer question directly with selected option or first provide reasoning and then which is known chain thought. In addition checking whether correct, evaluation look LLM-estimated probability its response...

This paper presents TELEIA, a dataset for the evaluation of Spanish language knowledge in Large Language Models (LLMs). TELEIA is designed to complement existing LLMs tests that evaluate many areas or tasks and are written English. To these English translated, which reasonable most technical tasks, but not when evaluating language. New specifically needed introduces an initial step direction. The as set multiple-choice questions have same format level those used several humans. enable...

In recent years, many proposals of context-aware systems applied to IoT-based smart environments have been presented in the literature. Most previous works provide a generic high-level structure how system can be operationalized, but do not offer clues on implement it. On other hand, there are implementations specific that context-specific: it is clear they extended use cases. this article, we aim an open-source reference implementation for providing data analytics capabilities environments....

The collaboration of the real world and virtual world, known as Digital Twin, has become a trend with numerous successful use cases. However, there are challenges mentioned in literature that must be addressed. One most important issues is difficulty Twins due to lack standardization their implementation. This article continues previous work proposed generic architecture based on FIWARE components build any field. Our proposes Linked Open Data mechanism facilitate communication Twins. We...

The introduction of Artificial Intelligence (AI) generative language models such as GPT (Generative Pre-trained Transformer) and tools ChatGPT has triggered a revolution that can transform how text is generated. This many implications, for example, AI-generated becomes significant fraction the text, would this have an effect on capabilities readers also training newer AI tools? Would it affect evolution languages? Focusing one specific aspect language: words; will use increase or reduce...

The term Data Space, understood as the secure exchange of data in distributed systems, ensuring openness, transparency, decentralization, sovereignty, and interoperability information, has gained importance during last years. However, Spaces are an initial phase definition, new research is necessary to address their requirements. Open ecosystem can be one precursors it provides mechanisms ensure information through resource discovery, exchange, aggregation via metadata. require more advanced...

The performance of conversational Large Language Models (LLMs) in general, and ChatGPT particular, is currently being evaluated on many different tasks, from logical reasoning or maths to answering questions a myriad topics. Instead, much less attention devoted the study linguistic features texts generated by these LLMs. This surprising since LLMs are models for language, understanding how they use language important. Indeed, poised have significant impact evolution languages as may...

The growing interest in Large Language Models (LLMs) and particular conversational models with which users can interact has led to the development of a large number open-source chat LLMs. These are evaluated on wide range benchmarks assess their capabilities answering questions or solving problems almost any possible topic test ability reason interpret texts. Instead, evaluation knowledge that these have languages received much less attention. For example, words they recognize use different...

Procrastination is one of the most common problems among students causing mental health issues, low motivation, and poor academic performance. The emergence Learning Management Systems has made it possible to accurately pinpoint when procrastination takes place due unobtrusive collection fine-grained data from students. However, studies that analyze regard as an intrinsic characteristic In this work, we study a process unfolds with time. We use sequence analysis map evolution students'...

The performance of conversational Large Language Models (LLMs) in general, and ChatGPT particular, is currently being evaluated on many different tasks, from logical reasoning or maths to answering questions a myriad topics. Instead, much less attention devoted the study linguistic features texts generated by these LLMs. This surprising since LLMs are models for language, understanding how they use language important. Indeed, poised have significant impact evolution languages as may...

Although many courses in computer science and software engineering require students to work on practical assignments, these are usually toy projects that do not come close real professional developments. As such, recent graduates often fail meet industry expectations when they first enter the workforce. In view of gap between graduates’ skills expectations, several institutions have resorted integrating open-source development as part their programs. this pilot study, we report results...

The introduction of ChatGPT has put Artificial Intelligence (AI) Natural Language Processing (NLP) in the spotlight. adoption been exponential with millions users experimenting it a myriad tasks and application domains impressive results. However, limitations suffers hallucinations, for example producing answers that look plausible but they are completely wrong. Evaluating performance similar AI tools is complex issue being explored from different perspectives. In this work, we contribute to...

Learning databases is indispensable for individuals studying software engineering, computer science, or involved in the IT industry. We have analyzed a remote educational escape room teaching four different higher education courses two consecutive academic years. employed three instruments evaluation: pre- and post-test to assess room's learning effectiveness, questionnaire gather students' perceptions, web platform that unobtrusively records interactions performance. show novel evidence...

The implementation of machine learning in Internet Things devices poses significant operational challenges due to limited energy and computation resources. In recent years, efforts have been made implement simplified ML models that can achieve reasonable performance while reducing energy, for example by pruning weights neural networks, or using reduced precision the parameters arithmetic operations. However, this type approach is implementation, i.e., loss accuracy model simplification....

We study the benefits and challenges of using Linked Open Data in smart city applications propose a set open source, highly scalable tools within case public-rental bicycle system, which can act as reference guide for other applications.

The wide adoption of Large language models (LLMs) makes their dependability a pressing concern. Detection errors is the first step to mitigating impact on system and thus, efficient error detection for LLMs an important issue. In many settings, LLM considered as black box with no access internal nodes; this prevents use schemes that need model's nodes. An interesting observation output in error-free operation should be valid normal text. Therefore, when text not or differs significantly from...