Self-Normalization Properties of Language Modeling

Jacob Goldberger, Oren Melamud


Abstract
Self-normalizing discriminative models approximate the normalized probability of a class without having to compute the partition function. In the context of language modeling, this property is particularly appealing as it may significantly reduce run-times due to large word vocabularies. In this study, we provide a comprehensive investigation of language modeling self-normalization. First, we theoretically analyze the inherent self-normalization properties of Noise Contrastive Estimation (NCE) language models. Then, we compare them empirically to softmax-based approaches, which are self-normalized using explicit regularization, and suggest a hybrid model with compelling properties. Finally, we uncover a surprising negative correlation between self-normalization and perplexity across the board, as well as some regularity in the observed errors, which may potentially be used for improving self-normalization algorithms in the future.
Anthology ID:
C18-1065
Volume:
Proceedings of the 27th International Conference on Computational Linguistics
Month:
August
Year:
2018
Address:
Santa Fe, New Mexico, USA
Editors:
Emily M. Bender, Leon Derczynski, Pierre Isabelle
Venue:
COLING
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
764–773
Language:
URL:
https://aclanthology.org/C18-1065
DOI:
Bibkey:
Cite (ACL):
Jacob Goldberger and Oren Melamud. 2018. Self-Normalization Properties of Language Modeling. In Proceedings of the 27th International Conference on Computational Linguistics, pages 764–773, Santa Fe, New Mexico, USA. Association for Computational Linguistics.
Cite (Informal):
Self-Normalization Properties of Language Modeling (Goldberger & Melamud, COLING 2018)
Copy Citation:
PDF:
https://aclanthology.org/C18-1065.pdf