How to automatically identify major research sponsors selecting keywords from the WoS Funding Agency field
In a context of increasingly limited resources, the demand for information from research funding bodies is growing. The exploitation of the funding acknowledgements collected in WoS publications can be useful for these sponsors, not only because it allows them to know the published results with thei...
| Autores: | , |
|---|---|
| Formato: | artículo |
| Estado: | Versión aceptada para publicación |
| Fecha de publicación: | 2018 |
| País: | España |
| Recursos: | Consejo Superior de Investigaciones Científicas (CSIC) |
| Repositorio: | DIGITAL.CSIC. Repositorio Institucional del CSIC |
| OAI Identifier: | oai:digital.csic.es:10261/172933 |
| Acesso em linha: | http://hdl.handle.net/10261/172933 |
| Access Level: | acceso abierto |
| Palavra-chave: | WoS Funding acknowledgements Funders identification Automatic procedures Performance evaluation Statistical analyses |
| Resumo: | In a context of increasingly limited resources, the demand for information from research funding bodies is growing. The exploitation of the funding acknowledgements collected in WoS publications can be useful for these sponsors, not only because it allows them to know the published results with their financial support, but also because it provides a framework to evaluate the efficiency of the different funding instruments. The present work adds to the knowledge of previous studies to offer a simple and efficient methodology that automatically identifies major sponsors, and their funded research, using keywords. To this end, articles with Spain in the address field and English in the language field are obtained (years 2010 2014), given that WoS only considers funding acknowledgements written in English. Subsequently, the Funding Agency (FA) field of these articles is treated, selecting funders' variants that will serve as keywords in the FTS (Full Text Search) for the location of the research supported by major sponsors. In addition, a sample of reviewed documents is provided to evaluate the reliability of the proposed methodology, performing also some statistical tests. The results show a recall of 91.5% of the sample articles, with a precision of 99%. Notwithstanding, there are differences in the automatic identification of funders by institutional sector and/or area, being the Government sector the one with the highest precision and recall, and the area of Agriculture, Biology & Environment the one with the best degree of association between the automatic classification and the reviewed one. Finally, possible future developments are offered, paying special attention to increasing the automation of the standardisation of funders' names. |
|---|