Abstract This paper is concerned with a class of fractional p(x, y)−Kirchhoff type problems with Dirichlet boundary data along with indefinite weight of the following form By means of direct variational approach and Ekeland’s variational principle, we investigate the existence of nontrivial weak solutions for the above problem in case of the competition between the growth rates of functions p and r involved in above problem, this fact is essential in describing the set of eigenvalues of this problem.

Resumen Este artículo estudia una clase de problemas de tipo p(x, y)−Kirchhoff fraccionarios con data Dirichlet en el borde junto con un peso indefinido de la siguiente forma A través del enfoque variacional directo y el principio variacional de Ekeland, investigamos la existencia de soluciones débiles no triviales para el problema anterior en el caso de competencia entre las tasas de crecimiento de las funciones p y r involucradas en el problema. Este hecho es esencial para describir el conjunto de valores propios de este problema.

ABSTRACT Sampling is an essential step in estimating a parameter: thus, cost and time associated to this step should be minimized. Sequential sampling is characterized by using samples of variable sizes given as a function of observations, and sequential sampling provides a smaller sample size than a fixed-size sample in most cases. In addition, the Bayesian decision theory can be incorporated into sequential sampling to perform parameter estimation because it allows the inclusion of a priori information about the parameter of interest, which optimizes the procedure. However, the great challenge to performing the Bayesian sequential estimation in establishing the stopping criteria. Most studies in this area investigate binomial distributions, while few analyze multinomial distributions. This study aimed to define the stopping criteria for the Bayesian sequential estimation of the parameters of multinomial distributions with conjugate Dirichlet priors. The proposed methodology was applied to a set of X-ray test data for quality control of maize seed lots. This test uses conventional sampling techniques in which a sample has a fixed size with 200 seeds. The influence of two priors on the stopping criteria was evaluated, one uniform and one conjugate, with hyperparameters based on reference information from the literature. The results indicated a reduction in the sample size in most lots evaluated. thus minimized observations fixedsize cases addition interest procedure However Xray X ray 20 seeds evaluated literature 2

Abstract At some moment in our lives, we are probably faced with the following question: How likely is it that you would recommend [company X] to a friend or colleague?. This question is related to the Net Promoter Score (NPS), a simple measure used by several companies as indicator of customer loyalty. Even though it is a well-known measure in the business world, studies that address the statistical properties or the sample size determination problem related to this measure are still scarce. We adopt a Bayesian approach to provide point and interval estimators for the NPS and discuss the determination of the sample size. Computational tools were implemented to use this methodology in practice. An illustrative example with data from financial services is also presented. lives company X colleague. colleague . colleague? NPS, , (NPS) loyalty wellknown well known world scarce practice presented (NPS

RESUMEN Este artículo pretende mapear cómo se ha desarrollado el campo de las capacidades dinámicas desde los artículos de Teece et al. (1997) y Eisenhardt y Martin (2000). Identificamos 10.838 artículos y utilizamos el algoritmo de asignación latente de Dirichlet para el modelado de temas. Realizamos dos tipos de análisis: el primero, basado en la temporalidad y características de los estudios; y el segundo, en la red de conceptos teóricos establecidos. Los resultados indican que existe aproximación entre las ideas de dos artículos seminales que inicialmente se veían como opuestas. Observamos un movimiento hacia la valoración de las cuestiones relacionales, siguiendo un camino de construcción colectiva y prestando menos atención a la empresa de forma aislada. El artículo contribuye a: señalar la consolidación del campo de las capacidades dinámicas; comprender los principales elementos involucrados en el desarrollo de las capacidades dinámicas; definir sus conceptos originales y actuales; e indicar las tendencias en el campo y una posible agenda para futuras investigaciones. al 1997 (1997 2000. 2000 . (2000) 10838 10 838 10.83 temas análisis primero estudios segundo establecidos opuestas relacionales aislada actuales investigaciones 199 (199 200 (2000 1083 1 83 10.8 19 (19 20 (200 108 8 10. (1 2 (20 ( (2

ABSTRACT This paper aims to map out how the field of dynamic capabilities has developed since the seminal studies by Teece et al. (1997) and Eisenhardt and Martin (2000). We identified 10,838 papers and used the Latent Dirichlet Allocation algorithm for topic modeling. We conducted two analyses: the first was based on the temporality and characteristics of the studies; the second was interpretative and based on the network of theoretical concepts. Results indicate an approximation between the ideas of the two seminal studies, which were initially viewed as opposite. We observe a movement to value relational issues, following a collective construction path, and paying less attention to the firm itself. Overall, we were able to understand the consolidation of the dynamic capabilities field; understand the core elements involved in the development of dynamic capabilities; set out the original and current concepts of dynamic capabilities; and indicate tendencies and a possible future research agenda. al 1997 (1997 2000. 2000 . (2000) 10838 10 838 10,83 modeling analyses opposite issues path itself Overall agenda 199 (199 200 (2000 1083 1 83 10,8 19 (19 20 (200 108 8 10, (1 2 (20 ( (2

RESUMO Este artigo visa mapear como o campo das capacidades dinâmicas se desenvolveu desde os artigos seminais de Teece et al. (1997) e Eisenhardt e Martin (2000). Identificamos 10.838 artigos e usamos o algoritmo de alocação latente de Dirichlet para modelagem de tópicos. Realizamos dois tipos distintos de análise: a primeira foi baseada na temporalidade e características dos estudos, enquanto a segunda foi interpretativa e baseada na rede de conceitos teóricos estabelecidos. Os resultados indicam que há uma aproximação entre as ideias dos dois artigos seminais, que inicialmente eram vistas como opostas. Observamos um movimento de valorização das questões relacionais, seguindo um caminho de construção coletiva e de menor atenção à empresa isoladamente. A contribuição do artigo está em apontar a consolidação do campo das capacidades dinâmicas e suas tendências, ampliar a compreensão dos principais elementos envolvidos no desenvolvimento dessas capacidades e definir seus conceitos originais e atuais, bem como oferecer uma possível agenda para futuras pesquisas. al 1997 (1997 2000. 2000 . (2000) 10838 10 838 10.83 tópicos análise estudos estabelecidos opostas relacionais isoladamente tendências atuais pesquisas 199 (199 200 (2000 1083 1 83 10.8 19 (19 20 (200 108 8 10. (1 2 (20 ( (2

Abstract Artificial intelligence tools are new in taphonomy and are growing fast. They are being used mainly to investigate bone surface marks. In order to investigate this subject, a bibliometric study was made to understand the growing rate of this intersectional field, the future, and gaps in the field until now. From Scopus and Google Scholar metadata, graphs were made to describe the data, and inferential statistics were made by regression with the Ordinary Least Squares method. Exploratory analysis with word clouds, topic modeling, and natural language processing with Latent Dirichlet Allocation as a method were also made using the entire corpus from the papers. From the first register until 2023, we found eight articles in Scopus and 32 in Google Scholar; the majority of the studies and the most cited were from Spain. The studies are growing fast from 2016 to 2018, and the regression shows that growth can be maintained in the coming years. Exploratory analysis shows the most frequent words are marks, models, data, and bone. Topic modeling shows that the studies are highly concentrated on similar problems and the tools to solve them, revealing that there is much more to explore with computational tools in taphonomy and paleontology as well. marks subject future now metadata data clouds papers 2023 3 Spain 201 2018 years models them well 202 20 2

Resumen Objetivo: Identificar de manera no supervisada mediante topic modeling los temas de mayor interés en el campo de la Salud Laboral y los Servicios de Atención a Domicilio de los artículos científicos publicados en la materia. Método: Este estudio empleó el algoritmo de Machine Learning no supervisado Asignación Latente de Dirichlet para el topic modeling y el lexicón NRC para la realización del análisis de sentimientos del corpus de las fichas documentales obtenidas de MEDLINE (vía PubMed) usando los descriptores “Salud Laboral” y “Servicios de Atención de Salud a Domicilio”. Resultados: Del total de 70 fichas documentales analizadas, se obtuvo que la intensidad de las emociones en los textos era baja (oscilando en valores de 5 a 10), teniendo una mayor representación los sentimientos positivos frente a los negativos en una relación de 60/40. No hubo una variación de las proporciones de las emociones con respecto al período del estudio. Se identificaron los cuatro temas de mayor interés en los artículos analizados: cuidado domiciliario y satisfacción de los cuidadores, período de lactancia, programas de rehabilitación, y actividad física para mitigación del dolor. Conclusiones: Se ha podido constatar que las metodologías del procesado de lenguaje natural pueden ser una gran herramienta de apoyo al análisis de artículos científicos. Concretamente, se ha logrado determinar de manera clara y no supervisada los temas de mayor interés en el campo de la Salud Laboral y la Atención de Salud a Domicilio.

Abstract Objective: To identify in an unsupervised manner through topic modeling the topics of greatest interest in the field of Occupational Health and Home Care Services from the scientific articles published on the subject. Method: The study used the unsupervised Machine Learning algorithm Dirichlet Latent Assignment for topic modeling and the NRC lexicon to carry out the sentiment analysis of the corpus of document files obtained from MEDLINE (via PubMed) using the descriptors “Occupational Health” and “Home Care Services”. Results: Of the total of 70 documentary files analyzed, it was obtained that the intensity of the emotions in the texts was low (ranging in values from 5 to 10), with positive feelings having a greater representation compared to negative ones in a ratio of 60/ 40. There was no variation in the proportions of emotions with respect to the study period. The four topics of greatest interest were identified in the articles analyzed: home care and caregiver satisfaction, breastfeeding period, rehabilitation programs, and physical activity to mitigate pain. Conclusions: It has been confirmed that natural language processing methodologies can be a great support tool for the analysis of scientific articles. Specifically, it has been possible to determine in a clear and unsupervised manner the topics of greatest interest in the field of Occupational Health and Home Care Services.

Resumen El campo de las redes sociales ha sufrido importantes transformaciones en los últimos veinticinco años, en particular con la introducción de aplicaciones y plataformas digitales, así como la incorporación de estudios de otros campos del conocimiento que adoptan el enfoque de redes sociales en sus análisis. Este artículo ofrece una visión general de la evolución de los tópicos de investigación en este ámbito entre 1997 y 2021 a partir de la modelización de temas. El estudio parte de la producción académica que se recupera de la base de datos Scopus, considerando ventanas temporales de un año y utilizando el software Mallet. Se obtienen siete temas, cuya evolución en el tiempo se describe. Se concluye que los temas relacionados con los medios de comunicación social, así como las redes sociales en línea son estudiados con especial intensidad en los últimos años.

Abstract The social network field has suffered significant transformations in the last 25 years, particularly with the introduction of social networks online, as well as incorporated studies from many other knowledge fields that adopt the social network approach in their analyses. This paper offers an overview of the evolution of research topics in this field between 1997 and 2021 based on topic modeling. The methodology used draws from the Scopus database, considering time windows of a year and using the software, Mallet. Seven topics are obtained, whose evolution over time is described. It is concluded that the topics related to social media and social networks online have been studied with special intensity in the last years.

Abstract In this work a 1D slightly doped n-type Silicon layer is considered. Irreversible thermodynamics transport equations are used to obtain the spatial particle distribution in isothermal stationary state. The excess particle transport is studied in low injection regime and quasi-neutrality condition. The material is subjected to Dirichlet-Neumann (N-D) conditions at the boundaries. We wonder if an asymmetry of the particle flux exists when the boundary conditions are inverted. We find that the asymmetry does exist and a rectification factor may reach the value 0.35. We conclude that particle flux rectification seems to be featuring the particle transport in slightly non-homogeneously doped semiconductor when the excess hole concentration is smaller than the equilibrium hole concentration.

ABSTRACT In this work, the solution of Poisson’s equation in a semi-disc under a Dirichlet boundary condition at the base and a Neumann boundary condition on the circumference is calculated. The solution is determined in terms of Green’s function, which is calculated in two ways, by the method of images and by solving its equation. In the particular case of Laplace’s equation, it is presented a second way to solve it, which uses separation of variables and a Fourier transform. work Poissons Poisson s semidisc semi disc Greens Green function ways Laplaces Laplace transform

Resumen Objetivo Bibliometría, cienciometría y otras metodologías relacionadas pueden ser utilizadas para descubrir las pautas básicas, conexiones con diferentes temas y áreas y elementos demográficos de un determinado asunto de la literatura. El objetivo de este estudio es descubrir las pautas y tendencias en la investigación de enfermería sobre fractura de cadera. Métodos Estudio exploratorio descriptivo clasificado en metodologías de bibliometría, cienciometría e inteligencia de negocios. La principal fuente de datos fue Web of Science. Se consideraron artículos relacionados publicados entre 1990 y 2020. En el análisis de datos abstractos, además de cuadros, análisis de texto y modelos de red utilizados en el mapeo bibliométrico, también se usó la Asignación Latente de Dirichlet ( Latent Dirichlet Allocation , LDA) enriquecida por inteligencia de negocios. Resultados Se identificaron 380 estudios sobre fractura de cadera (artículo=351, revisión=29). El número de artículos recopilados presentó una tendencia de crecimiento cada año, la mayor cantidad de artículos fue publicada en 2020. En términos de distribución por país de los 380 estudios, Estados Unidos (n=159), Suecia (n=52) y Australia (n=32) fueron los tres primeros países. Entre los diez temas identificados por la LDA, se destacaron cuidados de enfermería, mortalidad y rehabilitación. Enfermería, delirium y adulto mayor fueron las palabras clave más utilizadas. Fragilidad y rehabilitación fueron las palabras clave introducidas más recientemente. Conclusión Los temas cuidados de enfermería, rehabilitación y vivencias de cuidado fueron estudiados intensamente. Estudios sobre cuestiones que surgen con el aumento de la edad y donde los cuidados de enfermería son importantes pueden ser conducidos con el uso de inteligencia de negocios y bibliometría en un futuro. Bibliometría básicas literatura Science 199 2020 abstractos cuadros bibliométrico LDA 38 artículo=351, artículo351 artículo 351 (artículo=351 revisión=29. revisión29 revisión revisión=29 . 29 revisión=29) año n=159, n159 n n=159 159 (n=159) n=52 n52 52 (n=52 n=32 n32 32 (n=32 países Enfermería recientemente intensamente futuro 19 202 3 artículo=351 artículo35 35 (artículo=35 revisión2 revisión=2 2 n15 n=15 15 (n=159 n=5 n5 5 (n=5 n=3 n3 (n=3 1 20 artículo=35 artículo3 (artículo=3 revisión= n1 n=1 (n=15 n= (n= artículo=3 (artículo= (n=1 (n artículo= (artículo

Abstract Objective Bibliometrics, scientometrics and other related methodologies can be used to reveal the basic patterns, links to different subjects and areas, and demographic elements of a particular topic within the literature. The aim of this study is to reveal the patterns and trends in nursing research on hip fracture. Methods This descriptive exploratory study can be classified in methodologies of bibliometrics, scientometrics and business intelligence. The Web of Science was the main data source. Related articles published between 1990 and 2020 were considered. In addition to tables, text analytics, and network models used in bibliometric mapping, Latent Dirichlet Allocation (LDA) enriched by business intelligence was utilized in the analysis of abstract data. Results The total of 380 studies on hip fracture were identified (article=351, review=29). The number of retrieved articles exhibit an increasing trend by year, the highest number of articles was published in 2020. In terms of distribution of the 380 studies by country, the United States ( n =159), Sweden ( n =52) and Australia ( n =32) were the top three countries. Among the ten topics identified by LDA, nursing care, mortality, and rehabilitation were the prominent ones. Nursing, delirium, and elderly were the most frequently used keywords. Frailty and rehabilitation were the most recently introduced keywords. Conclusion The topics of nursing care, rehabilitation and care experiences have been studied intensively. Studies on issues that may occur with increasing age and where nursing care is important can be conducted in the future using business intelligence and bibliometric. Bibliometrics areas literature bibliometrics source 199 202 considered tables analytics mapping LDA (LDA 38 article=351, article351 article 351 (article=351 review=29. review29 review review=29 . 29 review=29) year country =159, 159 =159 , =159) =52 52 =32 32 countries mortality ones Nursing delirium keywords intensively 19 20 3 article=351 article35 35 (article=35 review2 review=2 2 15 =15 =5 5 =3 1 article=35 article3 (article=3 review= =1 = article=3 (article= article= (article

Resumo Objetivo Bibliometria, cienciometria e outras metodologias relacionadas podem ser usadas para revelar os padrões básicos, conexões com diferentes temas e áreas e elementos demográficos de um determinado tópico da literatura. O objetivo deste estudo é revelar os padrões e tendências na pesquisa de enfermagem sobre fratura de quadril. Métodos Estudo exploratório descritivo classificado em metodologias de bibliometria, cienciometria e inteligência de negócios. A Web of Science foi a principal fonte de dados. Foram considerados artigos relacionados publicados entre 1990 e 2020. Na análise de dados abstratos, além das tabelas, análise de texto e modelos de rede usados no mapeamento bibliométrico, também foi utilizada a Alocação Latente de Dirichilet ( Latent Dirichlet Allocation -LDA) enriquecida por inteligência de negócios. Resultados Foram identificados 380 estudos sobre fratura de quadril (artigo=351, revisão=29). O número de artigos levantados apresentou tendência de crescimento a cada ano, o maior número de artigos foi publicado em 2020. Em termos de distribuição dos 380 estudos por país, Estados Unidos (n=159), Suécia (n=52) e Austrália (n=32) foram os três primeiros países. Dentre os dez temas identificados pela LDA, destacam-se os cuidados de enfermagem, mortalidade e reabilitação. Enfermagem, delirium e idoso foram as palavras-chave mais utilizadas. Fragilidade e reabilitação foram as palavras-chave introduzidas mais recentemente. Conclusão Os temas cuidados de enfermagem, reabilitação e vivências de cuidado têm sido intensamente estudados. Estudos sobre questões que ocorrem com o aumento da idade e onde o cuidado de enfermagem é importante podem ser conduzidos futuramente com uso de inteligência de negócios e bibliometria. Bibliometria básicos literatura bibliometria 199 2020 abstratos tabelas bibliométrico LDA -LDA 38 artigo=351, artigo351 artigo 351 (artigo=351 revisão=29. revisão29 revisão revisão=29 . 29 revisão=29) ano país n=159, n159 n n=159 , 159 (n=159) n=52 n52 52 (n=52 n=32 n32 32 (n=32 países destacamse destacam se Enfermagem palavraschave palavras chave utilizadas recentemente estudados 19 202 3 artigo=351 artigo35 35 (artigo=35 revisão2 revisão=2 2 n15 n=15 15 (n=159 n=5 n5 5 (n=5 n=3 n3 (n=3 1 20 artigo=35 artigo3 (artigo=3 revisão= n1 n=1 (n=15 n= (n= artigo=3 (artigo= (n=1 (n artigo= (artigo

Abstract This paper presents a number of identities for Dirichlet series and series with Stirling numbers of the first kind. As coefficients for the Dirichlet series we use Cauchy numbers of the first and second kinds, hyperharmonic numbers, derangement numbers, binomial coefficients, central binomial coefficients, and Catalan numbers.

Resumen Este artículo presenta identidades para series de Dirichlet y series con números de Stirling de primera especie. Como coeficientes de las series de Dirichlet usamos números de Cauchy de primera y segunda especie, números hiperarmónicos, subfactoriales, coeficientes binomiales, coeficientes binomiales centrales y números de Catalan.

Abstract: The cultural industry assumed greater relevance as a productive system and expanded its market share with different forms of reception, transmission, and communication with the public, increasingly using the so-called classification and recommendation algorithms and manipulation of mass processed data, which do not require cyber-physical systems for cataloging andconstant feedback from all parties involved for cataloging. In this regard, this paper proposes a methodology to support the classification and creation of corresponding groups, automatically, of cultural productions of certain segments through Robot Process Automation (RPA) techniques, to first extract public data created by fans of certain cultural segments, and Latent Dirichlet Allocation (LDA), for the clustering of these productions based on the data of the terms extracted by RPA. As a case study for this proposal, we specifically observed the anime market, defined as an originally Japanese cultural product with high fan engagement and high annual production scale, supported by data obtained from two public databases data: MyAnimeList and AniDB, built collaboratively by fans. The application of the methodology allowed the automatic classification of anime, grouping them into topics that allow the proposal of a new demography of products of this genre in relation to the current one, providing a greater level of detail and allowing to contemplate the expansion of new themes.

Resumen: La industria cultural asumió mayor relevancia como sistema productivo y amplió su cuota de mercado con distintas formas de recepción, transmisión y comunicación con el público, con un creciente uso de los llamados algoritmos de clasificación, recomendación y manipulación de datos procesados en masa, que no requieren sistemas ciber-físicos para la catalogación ni una retroalimentación constante de todas las partes involucradas. En este sentido, este trabajo propone una metodología para apoyar la clasificación y creación de grupos correspondientes de forma automática de obras y producciones culturales de determinados segmentos mediante técnicas de Robot Process Automation (RPA) para extraer, primeramente, datos públicos creados por fans de determinados segmentos culturales, y Latent Dirichlet Allocation (LDA), para la agrupación de estos trabajos a partir de los datos de los términos extraídos por RPA. Como caso de estudio para se observó específicamente el mercado de animes, definido como un producto cultural originariamente japonés con un fuerte compromiso y apoyo de los fans y alta escala de producción anual, sustentado en datos obtenidos de dos bases de datos de carácter público construidos en colaboración por fans: MyAnimeList y AniDB. La aplicación de la metodología propuesta permitió la clasificación automática de animes, agrupándolos en temas que permiten proponer una nueva demografía de obras de este género diferente de la actual, proporcionando un mayor nivel de detalle y permitiendo contemplar la expansión de temas nuevos.

Resumo: A indústria cultural adquiriu maior relevância como sistema produtivo e ampliou sua participação no mercado com diversas formas de recepção, transmissão e comunicação com o público, com utilização cada vez maior dos chamados algoritmos de classificação, recomendação e manipulação de dados processados em massa, os quais prescindem de sistemas ciber-físicos para catalogação e retroalimentação constante por todas as partes envolvidas. Neste sentido, o presente trabalho propõe uma metodologia de apoio à classificação e criação de grupos correspondentes, de forma automática, de obras e produções culturais de determinados segmentos através de técnicas de Robot Process Automation (RPA), para primeiramente extrair dados públicos criados por fãs de certos segmentos culturais, e de Latent Dirichlet Allocation (LDA), para a agrupação destas obras com base nos dados dos termos extraídos por RPA. Como caso de estudo, observou-se em específico o mercado de anime, sendo este definido como um produto cultural originalmente japonês com alto engajamento e apoio dos fãs e alta escala de produção anual, tendo como suporte de dados aqueles obtidos em duas bases de dados públicas, o MyAnimeList e o AniDB, construídos de forma colaborativa por fãs. A aplicação da metodologia proposta permitiu a classificação automática de animes, agrupando-os em tópicos que possibilitam a proposição de uma nova demografia de obras deste gênero diferente em relação à atual, provendo um nível de detalhamento maior e permitindo contemplar a expansão de temáticas novas.

The narratives shared on social media during a health crisis such as COVID-19 reflect public perceptions of the crisis. This article provides findings from a study of the perceptions of South African citizens regarding the government's response to the COVID-19 pandemic from March to May 2020. The study analysed Twitter data from posts by government officials and the public in South Africa to measure the public's confidence in how the government was handling the pandemic. A third of the tweets dataset was labelled using valence aware dictionary and sentiment reasoner (VADER) lexicons, forming the training set for four classical machine-learning algorithms-logistic regression (LR), support vector machines (SVM), random forest (RF), and extreme gradient boosting (XGBoost)-that were employed for sentiment analysis. The effectiveness of these classifiers varied, with error rates of 17% for XGBoost, 14% for RF, and 7% for both SVM and LR. The best-performing algorithm (SVM) was subsequently used to label the remaining two-thirds of the tweet dataset. In addition, the study used, and evaluated the effectiveness of, two topic-modelling algorithms-latent dirichlet allocation (LDA) and non-negative matrix factorisation (NMF)-for classification of the most frequently occurring narratives in the Twitter data. The better-performing of these two algorithms, NMF, identified a prevalence of positive narratives in South African public sentiment towards the government's response to COVID-19.

Abstract We consider, for a bounded open domain Ω in Rn; (n ≥ 1) and a function u : Ω → ℝm; (m ≥ 1) the quasilinear elliptic system: (QESw)(f,g) (0.1) Which is a Dirichlet problem. Here, v belongs to the dual space , f and g satisfy some stan- dard continuity and growth conditions. we will show the existence of a weak solution of this problem in the four following cases: σ is mono- tonic, σ is strictly monotonic, σ is quasi montone and σ derives from a convex potential.

Resumen El objetivo de la investigación fue analizar el papel del aprendizaje automático de datos en las revisiones sistemáticas de literatura. Se aplicó la técnica de Procesamiento de Lenguaje Natural denominada modelado de tópicos, a un conjunto de títulos y resúmenes recopilados de la base de datos Scopus. Específicamente se utilizó la técnica de Asignación Latente de Dirichlet (LDA), a partir de la cual se lograron descubrir y comprender las temáticas subyacentes en la colección de documentos. Los resultados mostraron la utilidad de la técnica utilizada en la revisión exploratoria de literatura, al permitir agrupar los resultados por temáticas. Igualmente, se pudo identificar las áreas y actividades específicas donde más se ha aplicado el aprendizaje automático, en lo referente a revisiones de literatura. Se concluye que la técnica LDA es una estrategia fácil de utilizar y cuyos resultados permiten abordar una amplia colección de documentos de manera sistemática y coherente, reduciendo notablemente el tiempo de la revisión.

Abstract The objective of the research was to analyze the role of machine data learning in systematic literature reviews. The Natural Language Processing technique called topic modeling was applied to a set of titles and abstracts collected from the Scopus database. Specifically, the Latent Dirichlet Assignment (LDA) technique was used, from which it was possible to discover and understand the underlying themes in the collection of documents. The results showed the usefulness of the technique used in the exploratory literature review, by allowing the results to be grouped by theme. Likewise, it was possible to identify the specific areas and activities where machine learning has been applied the most, in relation to literature reviews. It is concluded that the LDA technique is an easy-to-use strategy and whose results allow a wide collection of documents to be approached in a systematic and coherent manner, notably reducing the review time.