Evaluating Embeddings using Syntax-based Classification Tasks as a Proxy for Parser Performance

Most of the data is available via the page for “What’s in an Embedding? Analyzing Word Embeddings through Multilingual Evaluation”, my EMNLP 2015 paper.

This paper adds the parser evaluation. The models for the parsers are quite big and therefore not online.

You can download the org-mode file which I used to collect & visualize all my data: evaluation.org