Uncertainty in Neural Network Word Embedding: Exploration of Threshold for Similarity
This addresses the reliability of word embeddings for downstream tasks like information retrieval, though it appears incremental as it builds on existing embedding methods.
The authors tackled the problem of determining when word embedding similarity scores genuinely indicate term relatedness by quantifying embedding uncertainty and introducing a general threshold to filter highly related terms. Their evaluation on four information retrieval collections showed results significantly better than baseline and equal to or indistinguishable from optimal results.
Word embedding, specially with its recent developments, promises a quantification of the similarity between terms. However, it is not clear to which extent this similarity value can be genuinely meaningful and useful for subsequent tasks. We explore how the similarity score obtained from the models is really indicative of term relatedness. We first observe and quantify the uncertainty factor of the word embedding models regarding to the similarity value. Based on this factor, we introduce a general threshold on various dimensions which effectively filters the highly related terms. Our evaluation on four information retrieval collections supports the effectiveness of our approach as the results of the introduced threshold are significantly better than the baseline while being equal to or statistically indistinguishable from the optimal results.