NEW MODELS FOR DATA AUGMENTATION IN TIME SERIES PROBLEMS
BASIC INFORMATION
Ph.D. Student: José de la Torre-López
Advisors: José Raúl Romero, Aurora Ramírez
Started on: December 2019
Keywords: machine learning, systematic literature review, bibliometrics
THESIS PROPOSAL
Conducting a systematic review of the scientific literature involves a manual work of collection, analysis, reading, filtering, and selection, which is arduous and prone to errors. A systematic literature review (SLR), usually extensive in length, consists of a compilation and review of articles, books and other research related to a given area of research. Hours of analysis are brought together in a single document by reviewing the most recent references and findings, as well as the most innovative techniques within the research area.
However, the elaboration of an SLR involves a great effort both in terms of time and resources. Do not overlook any possible source of information in the search phase is very complicated due to the diversity of types of sources that we find, including scientific databases such as Scopus or WoS (Web of Science) or publishers (IEEEXplore, ACM, ScienceDirect, etc.). It is also recommended to include the analysis of grey literature, i.e., publications in blogs and social networks with an impact in the research area.
Recently, the use of artificial intelligence (AI) techniques, such as machine learning (ML) or other possible approaches, are enabling the automation of many of the phases involved in the creation of a SLR. The long-term goal of these techniques is for the human work to be primarily verification or complementary to AI tasks. The current state of the art contemplates techniques that partially act on the process of creating an SLR and most of them focus mainly on text analysis. This limits and greatly conditions the result, so it would be interesting to delve into other interesting techniques such as the interaction between scientific communities or the evaluation of quality metrics and bibliometric factors of the publications to refine the results.
The main objective of the thesis is the elaboration of methods that allow the maximum automation of one or more phases of the development of an SLR through the use of ML techniques integrated in a transparent way for the final user, with the possibility of human intervention in the process of creation and/or verification of the SLR.
FUNDS
The development of this thesis is being supported by:
- Spanish Ministry of Science and Innovation and the European Regional Development Fund, under project PID2020-115832GB-I00
PUBLICATIONS ASSOCIATED WITH THIS THESIS
INTERNATIONAL JOURNALS
- J. De la Torre-López, A. Ramírez, J.R. Romero. “Artificial intelligence to automate the systematic review of scientific literature”. Computing. 2023.