International audience

We present a two-phase, data-driven framework to optimize photovoltaic (PV) maintenance in hotel networks. In Phase 1, we develop a site-specific regression model (mean $\mathbf{R}^{\mathbf{2}}=$ 0.98) to forecast soiling levels based on environmental and meteorological data. In Phase 2, these predictions were used in a linear optimization model to schedule cleaning actions that minimize total operational costs, including labor, equipment costs, and penalties from a shortfall in energy. This framework was applied to five hotels in Mallorca, Spain, where it adjusted dynamically to local conditions and resource limitations to support cost-effective maintenance decisions. The results showed an increase in energy production and a reduction in operational costs for the hotel sector. This work contributes to the disciplines of data science and resource management by combining data analysis and operational planning to further sustainable energy efforts in diverse regions.

This proposal aims to suggest improvements for the article titled “Low-Cost Data Centers in Developing Countries” with the aim of enhancing its content, relevance, and practicality. Our expertise in data infrastructure, we believe these improve-ments will provide valuable insights and guidance to readers interested in establishing efficient data centers in developing countries.

The proliferation of open knowledge graphs has led to a surge in scholarly research on the topic over the past decade. This paper presents a bibliometric analysis of the scholarly literature on open knowledge graphs published between 2013 and 2023. The study aims to identify the trends, patterns, and impact of research in this field, as well as the key topics and research questions that have emerged. The work uses bibliometric techniques to analyze a sample of 4445 scholarly articles retrieved from Scopus. The findings reveal an ever-increasing number of publications on open knowledge graphs published every year, particularly in developed countries (+50 per year). These outputs are published in highly-referred scholarly journals and conferences. The study identifies three main research themes: (1) knowledge graph construction and enrichment, (2) evaluation and reuse, and (3) fusion of knowledge graphs into NLP systems. Within these themes, the study identifies specific tasks that have received considerable attention, including entity linking, knowledge graph embedding, and graph neural networks.

In this letter, we explain how intuitive and explainable methods inspired from human physiology and computational biology can serve to simplify and ameliorate the way we process and generate knowledge resources.

This letter discusses the limitations of the use of filters to enhance the accuracy of the extraction of parenthetic abbreviations from scholarly publications and proposes the usage of the parentheses level count algorithm to efficiently extract entities between parentheses from raw texts as well as of machine learning-based supervised classification techniques for the identification of biomedical abbreviations to significantly reduce the removal of acronyms including disallowed punctuations.

Human similarity and relatedness judgements between concepts underlie most of cognitive capabilities, such as categorisation, memory, decision-making and reasoning. For this reason, the proposal of methods for the estimation of the degree of similarity and relatedness between words and concepts has been a very active line of research in the fields of artificial intelligence, information retrieval and natural language processing among others. Main approaches proposed in the literature can be categorised in two large families as follows: (1) Ontology-based semantic similarity Measures (OM) and (2) distributional measures whose most recent and successful methods are based on Word Embedding (WE) models. However, the lack of a deep analysis of both families of methods slows down the advance of this line of research and its applications. This work introduces the largest, reproducible and detailed experimental survey of OM measures and WE models reported in the literature which is based on the evaluation of both families of methods on a same software platform, with the aim of elucidating what is the state of the problem. We show that WE models which combine distributional and ontology-based information get the best results, and in addition, we show for the first time that a simple average of two best performing WE models with other ontology-based measures or WE models is able to improve the state of the art by a large margin. In addition, we provide a very detailed reproducibility protocol together with a collection of software tools and datasets as supplementary material to allow the exact replication of our results.

The ageing of the human population is a threat to many countries in the world and this fact creates new challenges for age-friendly living, recreational and working environments. Therefore, solutions that can support senior citizens (Longinos for the men and Longinas for the women) will be necessary, in order to help them stay actively involved in their professional life for longer. This is possible by designing fit for purpose working environments and by enabling flexible management of job-, leisure- and health-related activities, considering their needs at the workplace, at home and on the move, with a particular focus on fighting social isolation. This project presents a robotic digital solution that makes provision for Longinos/Longinas persons, that are above 70 years old, a single view of integrated health, business and social data spread respectively in online health communities, online project management websites and social networks, as well as the provision of a set of services, that will allow them to manage huge amounts of data.

Users recently are joining multiple online social networks simultaneously. The different accounts owned by the same user in multiple social networks are most of time isolated from each other. Identifying the same users, so called i-bridge, across networks is an important task for many interesting inter-network based applications such as viral marketing, presidential campaigns, product announcement, etc. In this paper, we tackle the problem of me edge identification across online social networks. Indeed, for each user in the source network, we extract the set of similar accounts in the target network and then a set of friends for each similar account in the target network will be extracted. A suitable comparison will be performed using similarity functions between the two sets of friends in the source and target networks to identify bridge users. Experiments are performed through the extraction of users from the two most popular social networks, Facebook and Twitter, and then extract the list of friends for these users. Then, the me edge link identification is built through the exploitation of the friends sets assigned to user accounts in different social networks. Experiments on two real-world social networks Facebook and Twitter provide a high identification rate of me edge links.

The inter-social networks data represent an important source of information for several research fields like the sentiment analysis, the content propagation and the determination of influential users. The user-centered graph of the social networks designs their connection through the users' profiles. It represents the flow of the contents propagation via inter-social networks. In this paper, we present a novel structure merging the user centered graphs of different socials networks. This structure allows the simulation and the visualization of such graphs illustrating the social networks. It is designed and developed as a plug-in within the known software Gephi1. It allows the definition of the graph structure including different parameters in relation with users and their relationships, and the generation of a graph which can be visualized and handled through several treatments present in Gephi.

Quantifying the semantic relation between words is a key element in several applications including the treatments at the meaning level. A great variety of approaches are proposed in order to quantify the semantic proximity between concepts or words. These approaches exploit computational models including the hierarchical and textual information of the semantic resources. Among these models, the distributional approaches quantify the semantic relations based on the co-occurrence information according to the target words. In this paper, we study the distributional semantics of three resources: the collaborative resources Wiktionary and Wikipedia, and the thesaurus WordNet through the word relatedness task. We exploit the glosses of WordNet and Wiktionary as a corpus formed by short and precise words, and the contents of Wikipedia articles. The experiments are performed using the known measures PMI and cosine, and a list of known benchmarks in semantic relatedness task. The results show that a small corpus formed by well formed sentences can lead to good correlations but limited coverage capacity. Despite the improvement in coverage capacity using Wikipedia, the correlations between human judgments and computed values do not follow the same enhancement degree.

The biomedical field is an especially relevant domain due to the continuous increase of biomedical textual resources. Recently, the computation of the similarity between concepts has been used to improve the performance of the information retrieval from biomedical sources. Moreover, information content has shown an important role in measuring semantic similarity of concepts. In this paper, we present an information content computing survey with a critical study. Then, we propose a new intrinsic information content computing method. This method attempts to improve the estimation of the semantic likeness between concepts. It exploits the structure of the semantic resource “MeSH”, a reference thesaurus in the biomedical field. Drawing on previous works, for a given concept, the number of its hypernyms, the number of its hyponyms, and, the depth of every hyponym have been considered in our new method. The latter was evaluated and compared with related works by referring to benchmarks widely used in the biomedical field. Experiment shows that our method is able to provide more accurate similarity evaluation and achieves significant performance improvements in terms of correlation coefficient compared to related works.

Computing Semantic Similarity (SS) between words is an important issue of many research fields. Several features from knowledge databases can be involved in the definition of semantic computing models. Gloss-based approaches exploit the short and precise description for a concept for better expressing its semantics. In this paper, we propose a new Gloss-based WordNet and Wiktionary semantic Similarity (G2WS) measure between words. In fact, the computational model is based on a noun factors computed using the glosses of the ancestors assigned to a specific concept within the WordNet “is a” taxonomy. The factorW of each noun is provided using an Information Content (IC) measure. The set of glosses assigned to each concept in WordNet are enriched by those from Wiktionary. Our work has been evaluated and compared with related works using a wide set of benchmarks conceived for word semantic similarity task. Obtained results show that the new similarity measure correlated better with human judgments than related works.