Un algoritmo novedoso para la detección de tareas repetitivas con secuencia lógica en el teclado
Cargando...
Fecha
2014
Autores
Londoño Gonzalez, Bairon
Título de la revista
ISSN de la revista
Título del volumen
Editor
Ediciones Universidad Simón Bolívar
Facultad de Ingenierías
Facultad de Ingenierías
Resumen
En este trabajo se desarrolla una herramienta para la detección de tareas repetitivas con secuencias lógicas realizadas a través de comandos del teclado, mediante el diseño e implementación de un algoritmo basado en el uso de autómatas finitos determinísticos y agentes de búsqueda de patrones. La novedad del algoritmo desarrollado radica en que está orientado a la detección de tareas repetitivas cuyas actividades tienen una secuencia lógica y que actualmente no se encuentran automatizadas por lo complejo que es esta labor.
El diseño del algoritmo partió de la clasificación de todos los comandos de Windows, luego de clasificar los comandos de Windows y tomar los que aplican a tareas repetitivas simples o de secuencia lógica en sus actividades se representaron en un autómata finito determinista con el fin de obtener una base de conocimiento de actividades que construyan tareas repetitivas, finalmente se construyó un Keylogger que capture los comandos del teclado y un Agente que se encarga de filtrar los comandos recibidos por el Keylogger, evaluar los comandos en el AFD y detectar tareas repetitivas.
El algoritmo diseñado se validó mediante un conjunto de pruebas realizadas sobre dos casos artificiales y dos casos reales, las cuales manifestaron un excelente desempeño del algoritmo dado que en todas las pruebas se detectó la tarea repetitiva en ejecución con un máximo de cinco actividades reales y una duración menor a un minuto.
In this thesis develops a tool for the detection of repetitive tasks with logical sequence performed through keyboard commands, through the design and implementation of an algorithm based on deterministic finite state machines using agents for search patterns. The novelty of the algorithm developed is that it is aimed at the detection of repetitive tasks whose activities have a logical sequence and that are not currently automated by how complex this work. The design of the algorithm was based on the classification of all Windows command after qualifying Windows commands and take that apply to simple repetitive tasks or logical sequence in their activities were represented in a deterministic finite automaton in order to obtain a knowledge base of activities that build repetitive tasks, finally a keylogger that captures keyboard commands and an Agent who is responsible for filtering the commands received by the keylogger, evaluate the commands in the AFD and detect repetitive tasks built. The proposed algorithm was validated by a set of tests on two artificial cases and two real cases, which showed an excellent performance of the algorithm since all tests repetitive running task was detected with a maximum of five actual activities and lasting less than a minute.
In this thesis develops a tool for the detection of repetitive tasks with logical sequence performed through keyboard commands, through the design and implementation of an algorithm based on deterministic finite state machines using agents for search patterns. The novelty of the algorithm developed is that it is aimed at the detection of repetitive tasks whose activities have a logical sequence and that are not currently automated by how complex this work. The design of the algorithm was based on the classification of all Windows command after qualifying Windows commands and take that apply to simple repetitive tasks or logical sequence in their activities were represented in a deterministic finite automaton in order to obtain a knowledge base of activities that build repetitive tasks, finally a keylogger that captures keyboard commands and an Agent who is responsible for filtering the commands received by the keylogger, evaluate the commands in the AFD and detect repetitive tasks built. The proposed algorithm was validated by a set of tests on two artificial cases and two real cases, which showed an excellent performance of the algorithm since all tests repetitive running task was detected with a maximum of five actual activities and lasting less than a minute.
Descripción
Palabras clave
Automatización de tareas repetitivas, Algoritmos para la detección de patrones, Comandos de teclado, Autómatas, Árboles de decisión, Automation of repetitive tasks, Algorithms for detecting patterns, Keyboard commands, Automata, Tree decision