Skip to main content

Main navigation

  • About ITEFI
  • Research
  • Formación y empleo
  • OpenLab
  • Servicios científico técnicos
  • Staff Directory

Improving LSTMs' under-performance in authorship attribution for short texts

authorship attribution
LSTM
Stylometry
Oliva. Christian; Palmero Muñoz, Santiago; Lago-Fernández, Luis F.; Arroyo Guardeño, David
Proceedings of the 2022 European Interdisciplinary Cybersecurity Conference
http://hdl.handle.net/10261/268091

We present a novel approach for conducting authorship attribution over tweets using Long-Short Term Memory networks (LSTMs). Vanilla LSTMs use the last hidden state for prediction. Our strategy introduces a mechanism based on Max Pooling to process all the hidden states simultaneously, which helps the model to better detect authors’ stylometry. We obtain a 4% accuracy improvement with respect to vanilla LSTMs.

ACKNOWLEDGEMENTS

This project has received funding from the European Union’s Hori zon 2020 Research and Innovation Programme under grant agreement No. 872855 (TRESCA project), as well as from Comunidad de Madrid (Spain) under the project CYNAMON (no. P2018/TCS- 4566), cofunded with FSE and FEDER EU funds, Spanish Government under project MINECO/FEDER PID2020-114867RB-I0, and Grant PLEC2021-007681 (project XAI-DisInfodemics) funded by MCIN/AEI/ 10.13039/501100011033 and by European Union NextGeneration EU/PRTR.

GiCSI

proyecto/s relacionado/s

  • IA explicable para desinformación y detección de conspiración durante infodemias. XAIDisInfodemics
    Plan Estatal de Investigación Científica y Técnica y de Innovación 2017-2020, Programa Estatal de I+D+i Orientada a los Retos de la Sociedad (AEI)
  • Trustworthy, Reliable and Engaging Scientific Communication Approaches. TRESCA
    Programa HORIZONTE'2020 (UE)
Acoustics and Non Destructive Evaluation (DAEND)
  • Environmental Acoustics (GAA)
  • G Carma: Materials Characterization by Non Destructive Evaluation
  • ULAB, Ultrasounds for Liquid Analysis and Bioengineering
Information and Communication Technologies (TIC)
  • Cybersecurity and Privacy Protection Research Group (GiCP)
  • Research group on Cryptology and Information Security (GiCSI)
    • Quantum Communications Laboratory (LCQE)
  • Multichannel Ultrasonic Signal Processing Group (MUSP)
Sensors and Ultrasonic Systems (DSSU)
  • Ultrasonic Systems and Technologies (USTG)
  • Nanosensors and Smart Systems (NoySi)
  • Ultrasonic Resonators for cavitation and micromanipulation (RESULT)
  • Advanced Sensor Technology (SENSAVAN)
  • Quantum Electronics (QE)
Laboratorios
  • Laboratorio de Acústica
  • Laboratorio de Metrología Ultrasónica Médica (LMUM)
  • Laboratorio de Comunicaciones Cuánticas
  • Laboratory for International Collaboration in Advanced Biophotonics Imaging

Instituto de Tecnologías Físicas y de la Información Leonardo Torres Quevedo  - ITEFI
C/ Serrano, 144. 28006 - Madrid • Tel.: (+34) 91 561 88 06  Contacto  •  Intranet
EDIFICIO PARCIALMENTE ACCESIBLE POR PERSONAS CON MOVILIDAD REDUCIDA