On the Processing and Analysis of Microtexts: From Normalization to Semantics † Yerai Doval 1

Author Contributions: Y.D. conceived, designed and performed the normalization experiments; Y.D. analyzed 
the data from the normalization experiments; D.V. conceived, designed and performed the sentiment analysis 
experiments; D.V. analyzed the data from the sentiment analysis experiments; Y.D. and D.V. wrote the paper.
Acknowledgments: Research partially funded by the Spanish Ministry of Economy, Industry and 
Competitiveness (MINECO) through projects FFI2014-51978-C2-2-R, TIN2017–85160–C2–1-R and TIN2017–
85160–C2–2-R; the Spanish State Secretariat for Research, Development and Innovation (which belongs to 
MINECO) and the European Social Fund (ESF) under a FPI fellowship (BES-2015-073768) associated to project 
FFI2014-51978-C2-1-R; and by the Galician Regional Government under project ED431D 2017/12. This research 
has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 
research and innovation programme (grant agreement No. 714150 – FASTPARSE), which covered the costs of 
open access publishing. We gratefully acknowledge NVIDIA Corporation for the donation of a GTX Titan X 
GPU used for this research.
Conflicts of Interest: The authors declare no conflict of interest. The founding sponsors had no role in the 
design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and 
in the decision to publish the results. 
References 
1. 
Doval, Y.; Vilares, J.; Gómez-Rodríguez, C. LYSGROUP: Adapting a Spanish microtext normalization 
system to English. In Proceedings of the Workshop on Noisy User-generated Text, Beijing, China, 31 July 
2015; pp. 99–105. 
2. 
Doval, Y.; Vilares, M.; Vilares, J. On the performance of phonetic algorithms in microtext normalization. 
ESWA 2018, 113, 213–222. 
3. 
Doval, Y.; Gómez-Rodríguez, C. Comparing Neural- and N-gram-based Language Models for Word 
Segmentation. JASIST