Improving transparency of the Colombian Peace Treaty with NLP

Francisco Barreras (), Mónica Ribero () and Felipe Suárez ()

No 15402, Documentos de Trabajo from Quantil

Abstract: Factorization methods and probabilistic models provide useful ways to represent text that can capture properties like sentence relevance, topics in text and even semantic similarity. In general, methods that yield a low-dimensional representation of large volumes of text have become more important and have gained attention in a diversity of fields since they have the potential of assisting in the understanding of vast ammounts of information. The Colombian Peace treaty documented an exhaustive list of agreements between the FARC guerilla and Colombian Armed Forces across six main sections spanning 297 pages. Surprisingly, the final version of the treaty was made public only for 40 days before Colombians had to vote for its approval in a plebiscite. Given that most of the general population probably would not read the document and the political environment (including the media) was highly polarized, there was a growing need for an unbiased and practical way to review the document before the vote. This paper describes the technical details behind the implementation of a tool that analyses the 2016 Colombian peace treaty. By combining Natural Language Processing techniques we were able to provide a web-service that helped increase transparency and unbiased reviewing of each section of the peace treaty.

Keywords: GloVe; Latent Dirichlet Allocation; Natural Language Processing; NMF; Peace Treaty. (search for similar items in EconPapers)
Pages: 7
Date: 2017-02-20
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://richter.quantil.co/wp-content/uploads/2017/ ... er-4_-Febrero-17.pdf
Our link check indicates that this URL is bad, the error code is: 500 Can't connect to richter.quantil.co:80 (No such host is known. )

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:col:000508:015402

Access Statistics for this paper

More papers in Documentos de Trabajo from Quantil
Bibliographic data for series maintained by Administrador ().