Search

Matrix factorization models are the core of current commercial collaborative filtering Recommender Systems. This paper tested six representative matrix factorization models, using four collaborative filtering datasets. Experiments have tested a variety of accuracy and beyond accuracy quality measures, including prediction, recommendation of ordered and unordered lists, novelty, and diversity. Results show each convenient matrix factorization model attending to their simplicity, the required prediction quality, the necessary recommendation quality, the desired recommendation novelty and diversity, the need to explain recommendations, the adequacy of assigning semantic interpretations to hidden factors, the advisability of recommending to groups of users, and the need to obtain reliability values. To ensure the reproducibility of the experiments, an open framework has been used, and the implementation code is provided.

Abstract Acute respiratory infections (ARIs) are a major cause of morbidity and mortality among children. The causative pathogens show geographic and seasonal variations. We retrospectively evaluated the frequency and seasonality of respiratory pathogens in children and adolescents (age: 0–19 years) with ARIs treated between January 1, 2021, and March 31, 2022, at a single center in Mexico. Out of 2400 patients, 1,603 were diagnosed with SARS-CoV-2 infection and 797 were diagnosed with other common respiratory pathogens (CRPs). Of the 797 patients, 632 were infected with one CRP and 165 with > 2 CRPs. Deaths occurred only in SARS-CoV-2-infected patients. Rhinovirus/Enterovirus, respiratory syncytial virus B, and parainfluenza virus 3 were the most prevalent in cases with single and multiple infections. CRP showed a high frequency between autumn and winter of 2021, with higher incidence of hospitalization compared to COVID-19. The main comorbidities were immunosuppression, cardiovascular disease (CD), and asthma. The frequency of CRPs showed a downward trend throughout the first half of 2021. CRPs increased in single- and co-infection cases between the fourth and fifth waves of COVID-19, probably due to decreased nonpharmaceutical interventions and changes in diagnostic tests. Age, cyanosis (symptom), and immunosuppression (comorbidity) were found to differentiate between SARS-CoV-2 infection and CRP infection.

One of the primary supports for extensive agriculture is pasture, which can suffer severe damage from insects including the Argentine stem weevil, Listronotus bonariensis. The main control method has been the infection of ryegrass with an endophyte fungus, forming a symbiotic association that produces alkaloids. In this study, we evaluated the impact of endophyte and peramine production on the weight of L. bonariensis across seven unnamed lines (LE161-LE167), and two Lolium perenne cultivars: Jumbo and Alto AR1. L. bonariensis adults fed on leaves from LE164, LE166, and ALTO AR1 showed weight losses of 13.3%, 17.1% and 18.2%, respectively. Similarly, the corresponding alkaloidal extract from LE164, LE166, and ALTO AR1 exhibited an antifeedant effect on L. bonariensis adults in laboratory assays, as observed through weight loss or low weight gain (−12.5%, 8.8% and 4.9%, respectively). Furthermore, one alkaloid, peramine, also elicited an antifeedant effect when incorporated into an artificial diet. Liquid chromatographic analysis of the alkaloid extract revealed that peramine was present in LE164, LE166 and ALTO AR1 in amounts ranging from 46.5–184.2 ng/g. Peramine was not detected in Jumbo and the remaining experimental lines. These data suggest that L. bonariensis were susceptible to peramine produced from endophyte infection in experimental lines LE164 and LE166, as well as ALTO AR1, affecting their feeding behavior.

La enfermedad SARS-Cov-2 (COVID-19) es responsable de una variedad de síntomas como fiebre, tos y fatiga, si bien es cierto que el padecimiento más frecuente se centra a nivel respiratorio y cardiovascular, existen reportes preliminares que han alertó sobre la posibilidad de que el SARS-Cov-2 pueda tener un potencial neuroinvasivo, ya que algunos pacientes han presentado síntomas neurológicos, entre ellos dolor de cabeza, mareos y en casos más severos eventos cerebrovasculares isquémicos, de igual manera ha descrito la aparición de anosmia y ageusia como síntomas tempranos de la infección por COVID-19. La literatura actual sugiere una asociación importante entre la infección por SARS-Cov-2 y el sistema nervioso central, por lo que es importante reconocer y controlar las complicaciones neurológicas potencialmente mortales secundarias a la infección por COVID-19

The relationship between phosphorus (P) availability and water restriction was explored in this study, focusing on its impact on phosphorus use efficiency (PUE) and water use efficiency (WUE) in various bromegrass (Bromus spp.) genotypes. Under controlled conditions, five bromegrass genotypes, as well as one ryegrass (Lolium perenne) cultivar, were compared by subjecting them to two P levels and two watering regimes. It was determined that combining water and phosphorus limitations led to reduced plant productivity. Initially, the ryegrass outperformed the bromegrass, but this result declined over time, while bromegrass exhibited consistent stability. Notably, under P and water stress, enhanced root development was observed in bromegrass compared to that in ryegrass. Distinct patterns of PUE and WUE allowed for the categorization of bromegrass genotypes into three groups. Genotype 3457 emerged as the most efficient, scoring 20 out of 24, while Pro 94-49 A achieved a score of only 10 out of 24. This study suggests that the drought resilience of bromegrass may be linked to increased root growth during the early vegetative stages, which potentially facilitates improved P acquisition. However, further validation through long-term field experiments is needed. The insights from this study are potentially valuable for use in shaping plant breeding programs by revealing the plant adaptation mechanisms for both P and water absorption.

Abstract Background Yearly, up to 1 million patients worldwide suffer from cutaneous leishmaniasis (CL). In Ecuador, CL affects an estimated 5000 patients annually. CL leads to reduced Health Related Quality of Life (HRQL) as a result of stigma in the Asian and Mediterranean contexts, but research is lacking for Ecuador. The objective of this study was to explore the influence of CL suspected lesions on the quality of life of patients in the Pacific and Amazon regions. Methods Patients for this study were included in the Amazonian Napo, Pastaza, and Morona Santiago provinces and the Pacific region of the Pichincha province. Participating centers offered free of charge CL treatment. All patients suspected of CL and referred for a cutaneous smear slide microscopy examination were eligible. This study applied the Skindex-29 questionnaire, a generic tool to measure HRQL in patients with skin diseases. All statistical analysis was done with SPSS Statistics version 28. Results The skindex-29 questionnaire was completed adequately by 279 patients who were included in this study. All patient groups from the Amazon scored significantly (P

Recommender Systems have become one of the most important tools for streaming and marketplace systems in recent years. Their increased use has revealed clear bias and unfairness against minorities and underrepresented groups. This paper seeks the origin of these biases and unfairness. To this end, it analyzes the demographic characteristics of a gold standard dataset and its prediction performance when used in a multitude of Recommender Systems. In addition, this paper proposes Soft Matrix Factorization (SoftMF), which tries to balance the predictions of different types of users to reduce the present inequality. The experimental results show that those biases and unfairness are not introduced by the different recommendation models and that they come from the socio-psychological and demographic characteristics of the used dataset.

BackgroundAn automated hand-hygiene monitoring system (AHHMS) was implemented in October 2019 at the Hospital Infantil de México Federico Gómez (HIMFG), a tertiary pediatric referral hospital, in four of the hospital wards with the highest rates of Healthcare Associated Infections (HAIs). The clinical and economic impact of this system had not yet been assessed prior to this study. This study aimed to evaluate if the AHHMS is a cost-effective alternative in reducing HAIs in the HIMFG.MethodologyA full cost-effectiveness economic assessment was carried out for the hospital. The alternatives assessed were AHHMS implementation vis-a-vis AHHMS non-implementation (historical tendency). The outcomes of interest were infection rate per 1,000 patient-days and cost savings as a result of prevented infections. Infection rate data per 1,000 patient-days (PD) were obtained from the hospital's Department of Epidemiology with respect to the AHHMS. As regards historical tendency, an infection-rate model was designed for the most recent 6-year period. Infection costs were obtained from a review of available literature on the subject, and the cost of the implemented AHHMS was provided by the hospital. The assessment period was 6 months. The incremental cost-effectiveness ratio was estimated. Costs are reported in US Dollars (2021). Univariate sensitivity and threshold analysis for different parameters was conducted.ResultsThe total estimated cost of the AHHMS alternative represented potential savings of $308,927–$546,795 US Dollars compared to non-implementation of the system (US$464,102 v. US$773,029–$1,010,898) for the period. AHHMS effectiveness was reflected in a diminished number of infections, 46–79 (−43.4–56.7%) compared to non-implementation (60 v. 106-139 infections).ConclusionThe AHHMS was found to be a cost-saving alternative for the HIMFG given its cost-effectiveness and lower cost vis-a-vis the alternate option. Accordingly, the recommendation was made of extending its use to other areas in the hospital.

In recent times, recommender systems (RSs) have been attracting a lot of attention from the research community because of their groundbreaking applications [...]

BackgroundAt the beginning of the current COVID-19 pandemic, it became critical to isolate all infected patients, regardless of their age. Isolating children has a negative effect on both, them and their parents/caregivers. Nevertheless isolation was mandatory because of the potential risk that visitation might have on COVID-19 dissemination mostly among health personnel.MethodsFrom the starting of the COVID-19 pandemic in our pediatric hospital visits were forbidden. This 2 months period (April–May) was called P1. In June parents were allowed to visit (P2), under a visiting protocol previously published. Hospital workers were monitored for the presence of COVID-19 symptoms and tested for the infection when clinically justified. The positivity proportion and the relative risk (RR) of COVID-19 among the health personnel between periods were calculated. The caregivers were also followed up by phone calls.ResultsSince April 2020 to November 2020, 2,884 health personnel were studied for 234 days, (318,146 workers days). Although the COVID-19/1,000 health personnel days rate decreased from one period to another (1.43 vs 1.23), no statistically significant differences were found. During P1, 16 patients with COVID-19 were treated. During the follow up none of the family members were infected/symptomatic in P1, while in P2, 6/129 (4.65%) were symptomatic or had a positive test. All of them initiated between 2 and 4 days after the patient's admission. As they also had some other infected family members it was not possible to ensure the source of infection. There were no statistically significant differences in the RR of COVID-19 in health personnel, (RR 1, 95% CI 0.69–1.06, p = 0.162).ConclusionsWhen safely implemented, allowing parents/caregivers to spend time with their hospitalized COVID-19 children does not increase the contagion risk for hospital workers or among themselves.

The lack of bias management in Recommender Systems leads to minority groups receiving unfair recommendations. Moreover, the trade-off between equity and precision makes it difficult to obtain recommendations that meet both criteria. Here we propose a Deep Learning based Collaborative Filtering algorithm that provides recommendations with an optimum balance between fairness and accuracy. Furthermore, in the recommendation stage, this balance does not require an initial knowledge of the users’ demographic information. The proposed architecture incorporates four abstraction levels: raw ratings and demographic information, minority indexes, accurate predictions, and fair recommendations. Last two levels use the classical Probabilistic Matrix Factorization (PMF) model to obtain users and items hidden factors, and a Multi-Layer Network (MLN) to combine those factors with a ‘fairness’ (ß) parameter. Several experiments have been conducted using two types of minority sets: gender and age. Experimental results show that it is possible to make fair recommendations without losing a significant proportion of accuracy.

Nowadays, it is mandatory to complement the traditional learning methods with active ones that enhance the student’s motivation and facilitate the development of technical and soft competences. LEGO®Serious Play is an experiential methodology designed to enhance innovation and performance in the business world and it is fully aligned with these needs. Previous researches show that this methodology has great potential in engineering education, and specifically in software engineering education. This paper presents an original LEGO®Serious Play activity to teach in a playful and active way software engineering concepts such as life cycle models and software development activities, which can be extrapolated to other engineering fields. This activity was validated through a case study involving 242 computer science students and it was supported by quantitative and qualitative data gathered from a survey and a post-test. The results indicate that the students found this activity highly fun and motivating as well as very useful to learn about the addressed topics and develop soft skills.

Health personnel (HP) have been universally recognized as especially susceptible to COVID-19. In Mexico, our home country, HP has one of the highest death rates from the disease. From the beginning of the SARS-CoV-2 pandemic, an office for initial attention for HP and a call center were established at a COVID-19 national reference pediatric hospital, aimed at early detection of COVID-19 cases and stopping local transmission. The detection and call center implementation and operation, and tracing methodology are described here. A total of 1,042 HP were evaluated, with 221 positive cases identified (7.7% of all HP currently working and 26% of the HP tested). Community contagion was most prevalent (46%), followed by other HP (27%), household (14%), and hospitalized patients (13%). Clusters and contact network analysis are discussed. This is one of the first reports that address the details of the implementation process of contact tracing in a pediatric hospital from the perspective of a hybrid hospital with COVID-19 and non-COVID-19 areas.

Abstract In this paper, we revisit the famous classical Samuelson’s multiplier–accelerator model for national economy. We reform this model into a singular discrete time system and study its solutions. The advantage of this study gives a better understanding of the structure of the model and more deep and elegant results.

This paper proposes a scalable and original classification-based deep neural architecture. Its collaborative filtering approach can be generalized to most of the existing recommender systems, since it just operates on the ratings dataset. The learning process is based on the binary relevant/non-relevant vote and the binary voted/non-voted item information. This data reduction provides a new level of abstraction and it makes possible to design the classification-based architecture. In addition to the original architecture, its prediction process has a novel approach: it does not need to make a large number of predictions to get recommendations. Instead to run forward the neural network for each prediction, our approach runs forward the neural network just once to get a set of probabilities in its categorical output layer. The proposed neural architecture has been tested by using the MovieLens and FilmTrust datasets. A state-of-the-art baseline that outperforms current competitive approaches has been used. Results show a competitive recommendation quality and an interesting quality improvement on large number of recommendations, consistent with the architecture design. The architecture originality makes it possible to address a broad range of future works.

Komatiites represent the oldest known terrestrial rocks, and their composition has been cataloged as the closest to that of the first terrestrial crust after the cooling of the magma ocean. These rocks could have been present in multiple environments on the early Earth and served as concentrators of organic molecules. In this study, the adsorption of five amino acids (glycine, lysine, histidine, arginine, and aspartic acid) on a natural komatiite, a simulated komatiite, and the minerals olivine, pyroxene, and plagioclase were analyzed under three different pH values: acid pH (5.5), natural pH of the aqueous solution of each amino acid and alkaline pH (11). Adsorption experiments were performed in solid–liquid suspensions and organic molecules were analyzed by spectrophotometry. The main objective of this essay was to determine if the complex surfaces could have participated as concentrators of amino acids in scenarios of the primitive Earth and if the adsorption responds to the change of charge of the molecules. The results showed that komatiite is capable of adsorbing amino acids in different amounts depending on the experimental conditions. In total, 75 systems were analyzed that show different adsorptions, which implies that different interactions are involved, particularly in relation to the type of amino acid, the type of solid material and the conditions of the medium.

Abstract This paper studies the equilibrium of an extended case of the classical Samuelson’s multiplier–accelerator model for national economy. This case has incorporated some kind of memory into the system. We assume that total consumption and private investment depend upon the national income values. Then, delayed difference equations of third order are employed to describe the model, while the respective solutions of third-order polynomial correspond to the typical observed business cycles of real economy. We focus on the case that the equilibrium is not unique and provide a method to obtain the optimal equilibrium.

Se describe la diatomea Hantzschia gadorensis como especie nueva a partir de muestras recogidas en una balsa ganadera en la Sierra de Gádor en Almería, España. Sus principales características son un bajo grado de dorsiventralidad, la ausencia de nódulo central y la presencia de conopeo que se registra por primera vez en el género, lo que permite distinguirla de especies similares. Los humedales de este sistema montañoso se confirman así como interesantes focos de biodiversidad aún por explorar.

As the use of recommender systems becomes generalized in society, the interest in varying the orientation of their recommendations is increasing. There are shilling attacks' strategies that introduce malicious profiles in collaborative filtering recommender systems in order to promote the own products or services or to discredit those of the competition. Academic research against shilling attacks has been focused in statistical approaches to detect the unusual patterns in user ratings. Nowadays, there is a growing research area focused on the design of robust machine learning methods to neutralize the malicious profiles inserted into the system. This paper proposes an innovative robust method, based on matrix factorization, to neutralize the shilling attacks. Our method obtains the reliability value associated with each prediction of a user to an item. By monitoring the unusual reliability variations in the items prediction, we can avoid promoting the shilling predictions to the erroneous recommendations. This paper openly provides more than 13 000 individual experiments involving a wide range of attack strategies, both push, and nuke, in order to test the proposed approach. The results show that the proposed method is able to neutralize most of the existing attacks; its performance only decreases in the not relevant situations: when the attack size is not large enough to effectively affect the recommendations provided by the system.

Recommender system (RS) clustering is an important issue, both for the improvement of the collaborative filtering (CF) accuracy and to obtain analytical information from their high sparse datasets. RS items and users usually share features belonging to different clusters, e.g., a musical-comedy movie. Soft clustering, therefore, is the CF clustering’s most natural approach. In this paper, we propose a new prediction approach for probabilistic soft clustering methods. In addition, we put to test a not traditional scientific documentation CF dataset: SD4AI, and we compare results with the MovieLens baseline. Not traditional CF datasets have challenging features, such as not regular rating frequency distributions, broad range of rating values, and a particularly high sparsity. The results show the suitability of using soft-clustering approaches, where their probabilistic overlapping parameters find optimum values when balanced hard/soft clustering is used. This paper opens some promising lines of research, such as RSs’ use in the scientific documentation field, the Internet of Things-based datasets processing, and design of new model-based soft clustering methods.

Recommender system is an information filtering tool used to alleviate information overload for users on the web. Collaborative filtering recommends items to users based on their historical rating information. There are two approaches: memory-based, which usually provides inaccurate but explainable recommendations; and model-based, whose recommendations are more precise but hard to understand. Here we propose a Bayesian model that not only provides us with recommendations as good as matrix factorization models, but these predictions can also be explained. The model is based on both user-based and item-based collaborative filtering approaches, which recommends items by using similar users' and items' information, respectively. Experiments carried out using four datasets present good results compared to several state-of-the-art baselines, achieving the best performance using the Normalized Discounted Cumulative Gain (nDCG) quality measure and also improving the prediction's accuracy in some datasets.

The Adapting Agriculture to Climate Change Project set out to improve the diversity, quantity, and accessibility of germplasm collections of crop wild relatives (CWR). Between 2013 and 2018, partners in 25 countries, heirs to the globetrotting legacy of Nikolai Vavilov, undertook seed collecting expeditions targeting CWR of 28 crops of global significance for agriculture. Here, we describe the implementation of the 25 national collecting programs and present the key results. A total of 4587 unique seed samples from at least 355 CWR taxa were collected, conserved ex situ, safety duplicated in national and international genebanks, and made available through the Multilateral System (MLS) of the International Treaty on Plant Genetic Resources for Food and Agriculture (Plant Treaty). Collections of CWR were made for all 28 targeted crops. Potato and eggplant were the most collected genepools, although the greatest number of primary genepool collections were made for rice. Overall, alfalfa, Bambara groundnut, grass pea and wheat were the genepools for which targets were best achieved. Several of the newly collected samples have already been used in pre-breeding programs to adapt crops to future challenges.

Visual representation of user and item relations is an important issue in recommender systems. This is a big data task that helps to understand the underlying structure of the information, and it can be used by company managers and technical staff. Current collaborative filtering machine learning models are designed to improve prediction accuracy, not to provide suitable visual representations of data. This paper proposes a deep learning model specifically designed to display the existing relations among users, items, and both users and items. Making use of representative datasets, we show that by setting small embedding sizes of users and items, the recommender system accuracy remains nearly unchanged; it opens the door to the use of bidimensional and three-dimensional representations of users and items. The proposed neural model incorporates variational embedding stages to “unpack” (extend) embedding representations, which facilitates identifying individual samples. It also replaces the join layers in current models with a Lambda Euclidean layer that better catches the space representation of samples. The results show numerical and visual improvements when the proposed model is used compared to the baselines. The proposed model can be used to explain recommendations and to represent demographic features (gender, age, etc.) of samples.

Premise Clearing leaves is a highly useful practice for many taxonomic, ecological, physiological, and eco‐physiological aspects of research. Using traditional methods, the procedure for clearing a leaf (referred to as diaphanization) can take several days or even weeks. In our laboratory we developed a technique, originally used for dissociating wood, that yields excellent epidermal and leaf venation preparations clearly showing the details of epidermal cells, hydathodes, trichomes, leaf margins, and leaf venation, in a maximum of three days. Methods and Results A solution originally used to macerate wood (Franklin’s solution) was used to accelerate the process of clearing leaves. Using this approach, it is possible to obtain clear images of the leaf adaxial and abaxial epidermal surfaces and leaf venation. Our technique works with both fragile and sturdy leaves, as well as with thin roots and stems. A few examples are presented. Conclusions Franklin’s solution is very simple to prepare and easy to manipulate. This new technique has the advantage of separating the leaf epidermal layers and venations, which can then be clearly observed.

En este artículo se aborda el análisis de los restos conservados del castillo de La Peña o fortaleza Alta de Martos (Jaén), importante yacimiento arqueológico que presenta estructuras emergentes de envergadura. Constituyó el castillo y sede principal del maestre de la Orden de Calatrava en la encomienda de Martos, territorio fronterizo con el reino nazarí de Granada entre los siglos XIII y XV, hasta que esta fortaleza fue abandonada en el siglo XVI. Además de la revisión crítica de la documentación existente sobre este singular ejemplo de arquitectura defensiva medieval, para realizar este análisis ha resultado fundamental el trabajo de campo y la elaboración de la planimetría completa y actualizada de las estructuras conservadas. Los resultados alcanzados en esta investigación arqueológica suponen un avance importante en el conocimiento de este Bien Patrimonial, y aportan nuevos datos que por ahora eran desconocidos o no se habían planteado. La documentación generada constituye la base para las intervenciones de restauración científica en este yacimiento.

This study aims at elucidating the environmental factors controlling benthic diatom diversity and uniqueness in Mediterranean mountain ponds. Samples of periphytic diatoms were collected in 45 ponds in Andalusia, south of Spain, and analysed by standard methods. Data analysis reveals that diatom diversity is mainly controlled by elevation and hydroperiod. Contrary to the usual findings in the literature, the highest scores on Shannon’s diversity index were found in high-elevation temporary ponds, but this effect is hidden by lake clustering in the analysed dataset. Significant distance-decay similarity (DDS) trends were detected in the analysis of floristic composition among the samples, stressing the importance of spatial factors that may override the effect of other abiotic factors. These findings highlight the role of isolation and dispersal limitation in the configuration of the biogeographical patterns of benthic diatoms.

Background In Mexico, coffee leaf rust (CLR) is the main disease that affects the Arabica coffee crop. In this study, the local response of two Mexican cultivars of Coffea arabica (Oro Azteca and Garnica) in the early stages of Hemileia vastatrix infection was evaluated. Methods We quantified the development of fungal structures in locally-infected leaf disks from both cultivars, using qRT-PCR to measure the relative expression of two pathogenesis recognition genes (CaNDR1 and CaNBS-LRR) and three genes associated with the salicylic acid (SA)-related pathway (CaNPR1, CaPR1, and CaPR5). Results Resistance of the cv. Oro Azteca was significantly higher than that of the cv. Garnica, with 8.2% and 53.3% haustorial detection, respectively. In addition, the non-race specific disease resistance gene (CaNDR1), a key gene for the pathogen recognition, as well as the genes associated with SA, CaNPR1, CaPR1, and CaPR5, presented an increased expression in response to infection by H. vastatrix in cv. Oro Azteca if comparing with cv. Garnica. Our results suggest that Oro Azteca’s defense mechanisms could involve early recognition of CLR by NDR1 and the subsequent activation of the SA signaling pathway.

Traditionally, recommender systems have been approached as regression models aiming to predict the score that a user would give to a particular item. In this work, we propose a recommender system that tackles the problem as a classification task instead of as a regression. The new model, Dirichlet Matrix Factorization (DirMF), provides not only a prediction but also its reliability, hence achieving a better balance between the quality and quantity of the predictions (i.e., reducing the prediction error by limiting the model’s coverage). The experimental results conducted show that the proposed model outperforms other models due to its ability to discard unreliable predictions. Compared to our previous model, which uses the same classification approach, DirMF shows a similar efficiency, outperforming the former on some of the datasets included in the experimental setup.

The efficiency of conventional heating energy source compared with Infrared (IR), Ultrasound (US), Microwave and the simultaneous combination US–IR eco-friendly approaches for preparation of new N-(5-R1 -amino-2-nitrophenyl)acetamides and 5-R1-amino-2-nitroaniline by Nucleophilic Aromatic Substitution (SNAr) via addition–elimination reactions on the halogens F, Cl, Br, I, employing amines as nucleophiles were explored. Moreover, phenyldiazenyl derivatives in good yields by an oxidative one-pot SNAr-based amination reaction from an unusual oxidation of 2-phenylhydrazinyl derivatives in DMSO was prepared.

Recommendation to a group of users is a big challenge for collaborative filtering. The recommendations to groups of users arise from the convenience of being able to recommend a group of users about products or services that satisfy the entire group. In this paper, we propose the similarity measure SMGU, tailored for collaborative filtering recommendations to groups of users. This similarity measure combines both numerical and non-numerical information. Numerical information is weighted attending to the rating singularity of the group members. This paper focuses on the assumption that the singularity of the ratings cast by the users of the group is relevant information for finding suitable neighbors. For each item, we consider that a rating is singular for a group or for a user when that rating is different from the majority of the rating cast by the other users. Non-numerical structural information can be considered as valuable to match group preferences with neighbors preferences. Experiments have been run using open recommender systems data sets. Compared with representative baselines, results show accuracy improvements when the proposed method is used. Additionally, this paper provides a section devoted to the experiments reproducibility issue. Finally, this paper opens opportunities to face new challenges in the recommendation to a group of users: explanation of recommendations, determination of reliability measures, and improvement of accuracy, novelty, and diversity results.

The existing scientific documentation-based recommender systems focus on exploiting the citations and references information included in each research paper and also the lists of co-authors. In this way, it can be addressed the recommendation of related papers and even related authors. The approach we propose is original because instead of using each paper citations and co-authors, we relate each of the papers with their main research topics. This approach provides a semantic level superior to that currently used, which allows us to obtain useful results. We can use collaborative filtering recommender systems to recommend research topics related to each paper and also to recommend papers related to each research topic. In order to face this innovative proposal, we have solved a series of challenges that allow us to offer various resources and results in the paper. Our main contributions are: 1) making a data mining of scientific documentation; 2) creating and publishing an open database containing the data mining results; 3) extracting the research topics from the available scientific documentation; 4) creating and publishing a recommender system data set obtained from the database and the research topics; 5) testing the data set through a complete set of collaborative filtering methods and quality measures; and 6) selecting and showing the best methods and results, obtained using the open data set, in the context of scientific documentation recommendations. Results of the paper show the suitability of the provided data set in collaborative filtering processes, as well as the superiority of the model-based methods to face scientific documentation recommendations.

Several collaborative filtering (CF) approaches have been developed in order to improve the quality of the recommendations. However, this improvement has always been measured as the average quality of the performed recommendations across all the users. It has not been analyzed for each individual user. In this paper, the existence of a more precise CF approach for each user is demonstrated. So, a novel hybrid method that merges recommendations provided by different CF approaches based on a multi-class classification algorithm is proposed. This classification is performed based on the user rating behavior. Experiments have been carried out on the MovieLens and Netflix datasets. The experimental results demonstrate an improvement on quality of both predictions and recommendations using the proposed hybrid CF approach. In addition, experiments have compared state-of-the-art baselines with the results obtained by the proposed approach.

In this paper, we propose a collaborative filtering method designed to improve the current memory-based prediction times without worsening and even improving the existing accuracy results. The accuracy improvement is achieved by combining the numerical relevance of the ratings with non-numerical information based on the votes structure. The improvement of the prediction time is achieved by setting four actions: 1) simplification of the similarity measure design, in order to minimize the necessary calculations; 2) construction and maintenance of a model that simplifies the predictions processing; 3) optimization of the computation, using a set-based model and a bit-based processing implementation; and 4) switching between the bit processing and the numerical processing, depending on the density of the users’ ratings. Experimental results show the improvements both in the prediction time and the accuracy. Experiments have used a significant amount of state-of-the-art baselines and collaborative filtering public data sets.

The use of resistant cultivars and fungicides are common methods to control coffee leaf rust (CLR), the main disease that affects the Arabica coffee crop. In this study, we evaluated the response of grafted and ungrafted plants during the early stage of Hemileia vastatrix infection. We used ungrafted plants of Oro Azteca (resistant cultivar) and Garnica (susceptible cultivar), and grafted plants, combining both as rootstock and graft (Garnica/Oro Azteca and Oro Azteca/Garnica). All plants were inoculated with H. vastatrix uredospores, and we quantified the development of fungal structures in the leaf tissue of inoculated plants using qRT-PCR to measure relative expression of two pathogenesis recognition genes (CaNDR1 and CaNBS-LRR) and three genes associated with the salicylic acid (SA) pathway (CaNPR1, CaPR1 and CaPR5). In Garnica grafted on Oro Azteca, the fungal structures recorded were significantly less than in Garnica ungrafted plants. In addition, the expression of defense-related genes in grafted plants was higher than in ungrafted plants. Our results indicate that the defense response to CLR is strongly influenced by the rootstock employed.

This study presents multiwall and bamboo-like carbon nanotubes found in samples from the Allende carbonaceous chondrite using high-resolution transmission electron microscopy (HRTEM). A highly disordered lattice observed in this material suggests the presence of chiral domains in it. Our results also show amorphous and poorly-graphitized carbon, nanodiamonds, and onion-like fullerenes. The presence of multiwall and bamboo-like carbon nanotubes have important implications for hypotheses that explain how a probable source of asymmetry in carbonaceous chondrites might have contributed to the enantiomeric excess in soluble organics under extraterrestrial scenarios. This is the first study proving the existence of carbon nanotubes in carbonaceous chondrites.

Pastures are important for the agricultural sector as forage, recreational and sports systems. The Chilean pasture seed market is highly dependent on introduced genetics; however, the effect of climate change and market conditions can modify pasture options. The objective of this study was to quantify changes of the pasture seeds sown by Chilean end-users, as metric tons (Mt) or proportion of the total (PT), based on 19 years of imported seed. The Mann–Kendall test and joinpoint regression analysis were used to estimate the overall trends and the average annual percent change (AAPC) for the whole period, respectively. The total imported Mt had an increasing trend and AAPC (+5.7%), wherein a large PT corresponded to ryegrasses (Lolium spp.) (0.681), with increasing trends for Mt and PT. Clovers (Trifolium spp.) had decreasing trends and AAPC (−2.9% for Mt and −9.6% for PT). For PT, the main species was perennial ryegrass (L. perenne L.) (0.357), with increasing trends for Mt. As Mt, a positive AAPC was found for plantain (P. lanceolata L.) (+17.4%) and chicory (C. intybus L.) (+63.2%). Over a period of 19 years, based on Mt but not PT, Chilean end-users have been adopting new pasture species with a decreasing use of Trifolium spp.

Red clover persistence has been one of the major targets for the most breeding programs worldwide. A phenotypic characterization at above- and belowground level was performed to a set of highly persistent red clover populations. The objective was to identify phenotypic changes occurring after empirical selection for plant survival, which is the consensus criteria of persistence. Eleven red clover populations were established on mesocosms of polyvinyl chloride (PVC) tubes of 11 cm in diameter and 100 cm depth, containing as substrate a mixture (v/v) of sand, vermiculite, soil and perlite. The trial was organized in a randomized complete block design with four replicates, each replicate consisting of five mesocosms per population and one plant per mesocosm. In total, 220 mesocosms were handled. At aboveground level, growth parameters, specific leaf area (SLA) and shoot dry matter (ShootDM) were measured. At belowground level, root morphology (volume, diameter and length) and topology (altitude, external path length [EPL] and dichotomous branching index [DBI]) were measured through image analyses. Analyses of variance were performed implementing a phenotypic linear mixed model using the Restricted Maximum Likelihood method. Additionally, variance components were estimated and broad-sense heritability was calculated for each phenotypic trait. Highly persistent cultivars exhibited 30% higher ShootDM and 10% lower leaf size and SLA than the oldest low-persistent cultivar Quiñequeli. At root level, they showed 20, 50 and 50% higher crown diameter, root length density and root volume than Quiñequeli, respectively, but 20% lower DBI. Root traits exhibited medium-low values of genetic control; broad sense heritability ranged between 0.20 and 0.48. In conclusion, highly persistent red clover cultivars and experimental lines bred in Chile modified their phenotypic expression of individual plants at shoot and root levels relative to the oldest low persistent cultivar Quiñequeli. Associations among above- and belowground traits offer opportunities for designing more efficient selection strategies. For instance, the strong relationship between SLA and root traits offers tremendous potential for indirect phenotypic selection.

Recommender Systems (RSs) have become an essential tool for the information society [...]

Internet of Things (IoT) projects are increasing in size over time, and some of them are growing to reach the whole world. Sensor arrays are deployed world-wide and their data is sent to the cloud, making use of the Internet. These huge networks can be used to improve the quality of life of the humanity by continuously monitoring many useful indicators, like the health of the users, the air quality or the population movements. Nevertheless, in this scalable context, a percentage of the sensor data readings can fail due to several reasons like sensor reliabilities, network quality of service or extreme weather conditions, among others. Moreover, sensors are not homogeneously replaced and readings from some areas can be more precise than others. In order to address this problem, in this paper we propose to use collaborative filtering techniques to predict missing readings, by making use of the whole set of collected data from the IoT network. State of the art recommender systems methods have been chosen to accomplish this task, and two real sensor array datasets and a synthetic dataset have been used to test this idea. Experiments have been carried out varying the percentage of failed sensors. Results show a good level of prediction accuracy which, as expected, decreases as the failure rate increases. Results also point out a failure rate threshold below which is better to make use of memory-based approaches, and above which is better to choose model-based methods.

Providing useful information to the users by recommending highly demanded products and services is a fundamental part of the business of many top tier companies. Recommender Systems make use of many sources of information to provide users with accurate predictions and novel recommendations of items. Here we propose, DeepMF, a novel collaborative filtering method that combines the Deep Learning paradigm with Matrix Factorization (MF) to improve the quality of both predictions and recommendations made to the user. Specifically, DeepMF performs successive refinements of a MF model with a layered architecture that uses the acquired knowledge in a layer as input for subsequent layers. Experimental results showed that the quality of both the predictions and recommendations of DeepMF overcome the baselines.

Depredation by the Iberian wolf (Canis lupus signatus) is currently thought to be a problem in some areas of Spain. However, there are few technically validated forensic tools available to determine the veracity of claims with a high degree of scientific confidence, which is important given that such attacks may lead to compensation. The analysis of bite marks on attacked animals could provide scientific evidence to help identify the offender. Thus, the aim of this study was to assess the morphological and morphometric characteristics of Iberian wolf dentition. This data collection would serve as a base-point for a more accurate identification of the wolves thorough their bite marks. For the first time, 36 dental variables have been studied in wolves’ skulls, employing univariate and multivariate analyses. The general morphological dental characteristics of wolves are very similar in terms of their dental formula and tooth structure to other canids, like domestic dogs. Sex differentiation was evident, principally in terms of the maxillary distance between the palatal surfaces of the canine teeth (UbC) and the width of the left mandibular canine teeth (LlCWd). New morphometric reference information was obtained that can aid the forensic identification of bite marks caused by the Iberian wolf with greater confidence.

Recommender systems aim to estimate the judgment or opinion that a user might offer to an item. Matrix-factorization-based collaborative filtering typifies both users and items as vectors of factors inferred from item rating patterns. This method finds latent structure in the data, assuming that observations lie close to a low-dimensional latent space. However, matrix factorizations have been traditionally designed by hand. Here, we present Evolutionary Matrix Factorization (EMF), an evolutionary approach that automatically generates matrix factorizations aimed at improving the performance of recommender systems. Initial experiments using this approach show that EMF generally outperforms baseline methods when applied to MovieLens and FilmTrust datasets, having a similar performance to those baselines on the worst cases. These results serve as an incentive to continue improving and studying the application of an evolutionary approach to collaborative filtering based on Matrix Factorization.