The Limitations of Cross-language Word Embeddings Evaluation
This work highlights a critical issue in natural language processing for researchers and practitioners, showing that current evaluation practices may be unreliable, making it incremental by exposing flaws in existing methods.
The paper investigates the limitations of cross-language word embeddings evaluation, finding that intrinsic and extrinsic evaluation methods lack correlation, as demonstrated by comparing 5 models on English-Russian datasets where scores on different intrinsic benchmarks did not correlate.
The aim of this work is to explore the possible limitations of existing methods of cross-language word embeddings evaluation, addressing the lack of correlation between intrinsic and extrinsic cross-language evaluation methods. To prove this hypothesis, we construct English-Russian datasets for extrinsic and intrinsic evaluation tasks and compare performances of 5 different cross-language models on them. The results say that the scores even on different intrinsic benchmarks do not correlate to each other. We can conclude that the use of human references as ground truth for cross-language word embeddings is not proper unless one does not understand how do native speakers process semantics in their cognition.