Maestría en Informática

Permanent URI for this collection

https://usc.scimago.es/handle/20.500.12421/67

Browse

Now showing 1 - 5 of 5

Arquitectura tecnológica para la asignación de ambulancias en ciudades inteligentes
(Universidad Santiago de Cali, 2020) Ospina Tobón, Luis Carlos; Dinas, Simena
The ambulances assignment to attend patients involved in accidents is one of the biggest problems in many third-world countries. It is a problem due mainly to two factors: mobility and limited resource availability. Traditionally, the problem is addressed by requesting the patient assistance by using radio broadcasting to all ambulance operators. The problem is that some notified ambulances may not be close to the accident, others could be on duty, on their way to attend other accidents, or not available. Nonetheless, it has also been notices that broadcasting generates competition among ambulances, or even lack of assistance to patients when many operators decide to attend the same accident. Furthermore, the competition to assist, has sometime generated new accidents. In other cases, negative impacts have been seen from the economic, ethic, social, politic, and mobility perspectives. Because of this situation, A Technological Architecture for the Assignment of Ambulances in Smart Cities was designed. Additionally, an intuitive and easy-to-use functional prototype was developed, applying the Design Thinking methodology. It is a proposal that puts technology at the service of community and aims to improve citizen quality of life, as proposed in the smart city manifest. This proposal was made under the cloud computational paradigm, which deploy the front-end of a mobile application for medical emergency real-time notifications. Additionally, the back-end manages the business logic to assign ambulances in a intelligent way and taking into account availability and georeferencing. The application was made possible by using technologies such as React JS, React Native, Firebase, Realtime Database, Cloud Functions, Google Maps, and Directions API. The software architecture is completely functional, scalable, and adequate. It has a TRL 4 maturity level, and complies with Jakob Nielsen’s 10 usability principles, giving a solution to the outlined needs. Tests was focused on Emergency logging and ambulance assignment, and in 98% of the cases, the prototype chose the nearest ambulance, the response was lower than 2 seconds, and the system behaved according to the necessary events to give solution to the problem.
Chatbots para apoyar la detección de síntomas de trastornos psiquiátricos
(Universidad Santiago de Cali, 2019) Cortés Loboa, John Edward; Dinas, Simena
This paper presents Juliana, a Chatbot prototype for Facebook Messenger that has the purpose of acting as an intelligent virtual assistant that allows the application of the MINI International Neuropsychiatric Interview (MINI). In addition to the application of the interview, Juliana has been given the possibility of notifying the result of the evaluation, indicating whether or not a specific user may have presumed symptoms of Major Depressive Episode and Risk of Suicide disorders. For doing so, models capable of understanding user’s queries in Spanish language using the Natural Language Processing skills of IBM Watson Assistant service were generated. Finally, in order to evaluate what the user experience is like when using Juliana, a quantitative type and descriptive subtype experimentation scenario was carried out with a sample of 50 users. Results show that Juliana offers a good user experience. However, it is important to consider those users who say they disagree with the quality of the system, the information and interface since improvements can be made to achieve a better experience for everyone.
Modelo multiclasificador aplicando minería de datos para el diagnóstico médico utilizando datos abiertos
(Universidad Santiago de Cali, 2019) Bolaños Ramírez, Harold Ferney; Dinas, Simena
In order to understand, analyze and eventually make use of a massive amount of data, a multidisciplinary approach has been proposed: data mining, it consists in the process of identifying patterns from massive databases. Data mining can be used in several fields: mobile computing, web mining, financial analysis, crime analysis, engineering, management and medicine. In medicine, data mining techniques can be used by researchers for the diagnosis and prediction of sundry diseases. Researchers have widely used several statistical analysis and data mining techniques to improve the accuracy in the diagnosis of the different diseases in medical care. The use of individual data mining techniques reveals an acceptable level of accuracy to determinate the diagnosis of the disease. Sets or combinations of classifiers have produced promising results, improving accuracy, reliability and, most importantly, extensive coverage in many different and practical applications. In this study, we evaluated the application of feature selection methods and the use of a set of three classifiers based on three different mixing strategies using open diabetes and breast cancer databases contained in the ICU data repositorio with the goal of supporting decision-making in medical diagnosis. The effectiveness of the combinations was measured using three commonly accepted metrics, as well as the response times in the construction of the models. Analysis of the results reveal that the sensitivity, specificity and percentage of classification are higher compared to the individual techniques and provides the basis for establishing an improved multi-classifier model.
Prototipo web para el modelo de acompañamiento y seguimiento a estudiantes de la UNIAJC
(Universidad Santiago de Cali, 2019) Oviedo Bayan, Gustavo Adolfo; Dinas, Simena
The Ministry of National Education (MEN) believes that dropping out of higher education for students or dropping out of higher education has become a phenomenon that affects institutions in Colombia. Each student faces social, academic and financial problems, which ultimately replicate in universities, since students are the raison d'être of a university, so it is necessary to implement a model to try to mitigate the phenomenon. The Information and Communication Technologies can support with information to help solve this problem, but not all the technological developments contribute what is necessary for the Institutions of higher education, as it is presented by the investigations that are reported in the literature of this article, an investigation was made with different universities at national level and consultations on the aspects that contribute to the desertion at international level. The creation of a model of characterization and follow-up to students, supported in a WEB prototype for the Antonio José Camacho University Institution, considering that this model can be adapted and implemented in any Higher Education Institution.
Sistema de información para la articulación y administración de datos dinámicos del sector educativo (SIADSE) en el departamento del Cauca
(Universidad Santiago de Cali, 2019) Bambague Meneses, Rodrigo Gaitán; Sánchez Ledesma, Javier Arley; Dinas, Simena
Nowadays, missional entities, in a public and private educational field, provide heterogeneous information systems (IS) to its service and use, delaying articulating the information. The spread of architecture and missional differences hinder the decision making and negatively impact the education fundamental right in the Cauca department. There by designing and developing a IS is proposed to articulate and analyze dynamically the different data sources information to facilitate the strategic educational field making decisions, following a protocol to an analysis and articulation data section. In line with the Ministry of national education normativity, this protocol is used to place strategical value to the information. This technological applied research proposes developing the objective, creating a single data source based on the analytics of the BIG DATA concepts. As a result, it seeks to generate reports or managerial sights that support the decisions agreed with the necessities, facilitating the optimal allocation of resources to the different educational programs and projects. All that is based on the SI development rooted in the XP and APEX methodology as an RAP development tool where they dynamically interacted to develop SIADSE.

Browse

Browsing Maestría en Informática by Author "Dinas, Simena"

Results Per Page

Sort Options