Verb sense induction with tensor decomposition

Márton Makrai

March 22, 2019, 14:00
SZTAKI, Lágymányosi u. 11.

I submitted a paper to ACL main conf with the following abstract, and plan to submit it to workshops as well. The main question of the work in progress talk is now how to evaluate the decomposition quantitatively.

We present verb representations based on dependency relations and tensor-decomposition applying the rather recent Orth-ALS decomposition algorithm to a subject × preverb × verb × object tensor in Hungarian, where preverbs (verb-modifying particles) interfere with argument structure and polisemy. The work-in-progress paper evaluates the obtained representations qualitatively, by clustering the verb embedding vectors, and showing some which intuitively fit to some semantic verb class.

Feb. 27, 2019	UMAP continued	Gábor Borbély
Feb. 1, 2019	UMAP	Gábor Borbély
Dec. 7, 2018	Deep contextualized word representations	Gábor Berend
Oct. 26, 2018	Approaches to Surface Realization of Universal Dependencies	Gábor Recski
Oct. 19, 2018	The Book of Why	Péter Földiák
Oct. 12, 2018	Borbély: Wasserstein VAE, Makrai: mondatklaszterezés	Gábor Borbély
Sept. 28, 2018	Making chatbots better by entropy-based data filtering (and future directions)	Richárd Csáky
Sept. 21, 2018	SkipGram - Zipf + Uniform = Vector Additivity	Gábor Borbély
Aug. 22, 2018	Expanding Access to Language through Data-Driven Design	Danielle Bragg
April 13, 2018	Semantic compression and rate distortion theory for memory representations	Dávid Gergely Nagy
Feb. 9, 2018	Bayesian model comparison of distributions with unequal supports	Gábor Borbély
Dec. 14, 2017	Comprehensive Conditioning of Neural Conversational Models	Richárd Csáky
Nov. 30, 2017	Bayesian model selection, inference and Minimum Description Length	Péter Földiák
Nov. 23, 2017	Hungarian morphological segmentation using recurrent and convolutional networks	Judit Ács
Nov. 9, 2017	The length of natural sentences	Gábor Borbély
Nov. 2, 2017	Brainstorming for SemEval Task 10: Capturing Discriminative Attributes	Márton Makrai
Sept. 21, 2017	Recurrent dropout	Dávid Márk Nemeskey
Sept. 14, 2017	Stanford’s Graph-based Neural Dependency Parser	Gábor Recski
Sept. 7, 2017	Chatbots	Gábor Recski
June 15, 2017	Sparse Coding of Neural Word Embeddings for Multilingual Sequence Labeling	Márton Makrai
June 1, 2017	Wasserstein GAN	Gábor Borbély
May 11, 2017	Autoencoder experiments on Hungarian words	Judit Ács
April 27, 2017	Beyond RNN: multi-dimensional RNN, RNN transducers, RNN grammars	Dávid Márk Nemeskey
April 13, 2017	Training a Universal Word Embedding	Eszter Iklódi
March 30, 2017	Interpreted Regular Tree Grammars for semantic parsing	Gábor Recski
March 16, 2017	2 recent papers on Deep Learning
March 2, 2017	Visualizing and Understanding Recurrent Networks	Gábor Borbély