Contributions to Proceedings:
K. A. Gemes, A. Kovacs, G. Recski:
"Machine comprehension using semantic graphs";
in: "Proceedings of the Automation and Applied Computer Science Workshop 2019",
AACS,
2019.
English abstract:
We present a novel method for recognizing entailment using semantic
graphs and apply it to the 2018 Semeval task on Machine Comprehension (MC). Concept graphs are built automatically from MC texts, questions, and answers, using the 4lang system. A strong baseline method using only these graphs is presented, followed by an enhancement of a state-of-the-art system. Preliminary results suggest that these features achieve a .5 percentage point improvement over the original system.
Keywords:
Semantic parsing; Natural language processing; 4lang; Graph transformation; Comprehension
Electronic version of the publication:
https://hlt.bme.hu/media/pdf/AACS19_paper_11.pdf
Created from the Publication Database of the Vienna University of Technology.