Publication: Linked Data para la generación de conocimiento financiero a partir de la extracción de información semiestructurada
Loading...
Identifiers
Publication date
2014-11
Defense date
2015-02-24
Authors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
En la actualidad, la información es generada por datos ubicados en un entorno distribuido pero vinculado. Con relación a esta premisa, las tecnologías semánticas y Linked Data, proporcionan un paradigma en el que no sólo los documentos, sino que también los datos son recursos de primera clase en la Web, permitiendo su extensión y la compartición de conocimientos hacia un espacio global de datos basado en estándares abiertos, mejor conocido como Web de datos.
En este trabajo de tesis, se presenta un modelo semántico inspirado en los principios de Linked Data que ofrece una alternativa de solución a los problemas de integración de datos que se manifiestan en los Estados financieros publicados por las empresas bajo el estándar XBRL a través de la Web. En este sentido, mediante el modelo semántico se identifican y subsanan ciertas limitaciones existentes en las Hojas de Balance, Cuentas de resultados y Estados de flujos de efectivo. Entre estas limitaciones destacan la falta de una semántica que permita la integración de sus datos para hacerlos navegables, la dificultad para el acceso a los mismos a través de protocolos asociados a Internet como el HTTP para la navegación e interconexión con otras fuentes de información, y la carente capacidad para la búsqueda de ratios financieros, así como el procesamiento de cálculos permitan un análisis fundamental o clásico que sirva de apoyo a la toma de decisiones.
El modelo semántico, se integra de taxonomías financieras basadas en la norma US-GAAP, y es el soporte fundamental de una base de conocimientos financieros reutilizable inspirada en Linked Data. En relación con lo anterior, la investigación que se realiza en este trabajo de tesis se sintetiza a través de la solución que se proporciona a las hipótesis que en ella se plantean, y mediante la que se busca demostrar que el modelo semántico tiene la capacidad para poblar una base de conocimientos financieros a partir de la integración de fuentes de datos externas, facilitar la reutilización de sus datos con terceros a través de Linked Data, ayudar a mejorar la calidad estructural de los datos financieros y comprobar que este modelo también facilita el análisis fundamental financiero para apoyar la toma de decisiones tanto automatizada como por parte de las personas.
At present, the information is generated by data located in a distributed environment but linked. In relation to this premise, semantic and Linked Data technologies provide a paradigm in which not only documents but also the data are first-class resources on the Web, allowing its extension and sharing of knowledge towards a global space data based on open standards, better known as Web of data. In this thesis, a semantic model based on the principles of Linked Data that provides an alternative solution to the problems presented data integration that are manifested in the financial statements published by the company under the XBRL standard through Web. In this sense, using the semantic model are identified and remedied certain limitations in Balance Sheets, Income Statements and Cash Flow Statements. These limitations include the lack of a semantics that allow the integration of their data to make it navigable, difficulty accessing them through associated Internet protocols like HTTP for navigation and interconnection with other sources of information and lacking ability for search of financial ratios, as well as processing calculations that allow a fundamental or classical financial analysis that supports the decision making. The semantic model integrated financial taxonomies based on US-GAAP standard, and is the main support base of reusable financial knowledge inspired Linked Data. In connection with this, the research conducted in this thesis is synthesized through the solution provided to the hypothesis raised therein, and by which seeks to demonstrate that the semantic model has the ability to financial populate a knowledge base from the integration of external data sources, facilitating the reuse of data with third parties via Linked data, help improve the structural quality of financial data and verify that this model also facilitates analysis financial crucial to support decision-making both automated and by the people.
At present, the information is generated by data located in a distributed environment but linked. In relation to this premise, semantic and Linked Data technologies provide a paradigm in which not only documents but also the data are first-class resources on the Web, allowing its extension and sharing of knowledge towards a global space data based on open standards, better known as Web of data. In this thesis, a semantic model based on the principles of Linked Data that provides an alternative solution to the problems presented data integration that are manifested in the financial statements published by the company under the XBRL standard through Web. In this sense, using the semantic model are identified and remedied certain limitations in Balance Sheets, Income Statements and Cash Flow Statements. These limitations include the lack of a semantics that allow the integration of their data to make it navigable, difficulty accessing them through associated Internet protocols like HTTP for navigation and interconnection with other sources of information and lacking ability for search of financial ratios, as well as processing calculations that allow a fundamental or classical financial analysis that supports the decision making. The semantic model integrated financial taxonomies based on US-GAAP standard, and is the main support base of reusable financial knowledge inspired Linked Data. In connection with this, the research conducted in this thesis is synthesized through the solution provided to the hypothesis raised therein, and by which seeks to demonstrate that the semantic model has the ability to financial populate a knowledge base from the integration of external data sources, facilitating the reuse of data with third parties via Linked data, help improve the structural quality of financial data and verify that this model also facilitates analysis financial crucial to support decision-making both automated and by the people.
Description
Keywords
Integración de datos, Linked data, Web de datos, Datos financieros, Empresas, XBRL