Multilingual LAMA: Investigating Knowledge in Multilingual Pretrained Language Models
This work addresses the lack of diversity in knowledge base research by extending it to multiple languages, though it is incremental as it builds on existing monolingual benchmarks.
The study investigated whether multilingual BERT (mBERT) can serve as a knowledge base across 53 languages, finding that performance varies by language and pooling predictions improves results, but mBERT shows language biases, such as predicting Italy for queries in Italian.
Recently, it has been found that monolingual English language models can be used as knowledge bases. Instead of structural knowledge base queries, masked sentences such as "Paris is the capital of [MASK]" are used as probes. We translate the established benchmarks TREx and GoogleRE into 53 languages. Working with mBERT, we investigate three questions. (i) Can mBERT be used as a multilingual knowledge base? Most prior work only considers English. Extending research to multiple languages is important for diversity and accessibility. (ii) Is mBERT's performance as knowledge base language-independent or does it vary from language to language? (iii) A multilingual model is trained on more text, e.g., mBERT is trained on 104 Wikipedias. Can mBERT leverage this for better performance? We find that using mBERT as a knowledge base yields varying performance across languages and pooling predictions across languages improves performance. Conversely, mBERT exhibits a language bias; e.g., when queried in Italian, it tends to predict Italy as the country of origin.