Semantic sentence similarity: size does not always matter
This work challenges the trend of using larger datasets in machine learning by showing that database characteristics like paraphrasing can be more important for semantic similarity tasks.
The study investigated whether visually grounded speech recognition models can learn sentence semantics without prior linguistic knowledge, finding that a model trained on a small image-caption database outperformed models trained on larger databases, with results correlating well with human judgments.
This study addresses the question whether visually grounded speech recognition (VGS) models learn to capture sentence semantics without access to any prior linguistic knowledge. We produce synthetic and natural spoken versions of a well known semantic textual similarity database and show that our VGS model produces embeddings that correlate well with human semantic similarity judgements. Our results show that a model trained on a small image-caption database outperforms two models trained on much larger databases, indicating that database size is not all that matters. We also investigate the importance of having multiple captions per image and find that this is indeed helpful even if the total number of images is lower, suggesting that paraphrasing is a valuable learning signal. While the general trend in the field is to create ever larger datasets to train models on, our findings indicate other characteristics of the database can just as important important.