Efficient comparison of sentence embeddings
Spyros Zoupanos, Stratis Kolovos, Athanasios Kanavos, Orestis Papadimitriou, Manolis Μaragoudakis
Abstract
The evolution of natural language processing (NLP) has drastically improved numerous applications in terms of quality of results and speed, like the use of semantic search in modern search engines. NLP has highly benefited from the recent developments in word and sentence embeddings which enable the transformation of complex NLP tasks, such as semantic similarity or Question and Answering (Q&A), into much simpler to perform vector comparisons. However, the new problems resulting from such transformations have also challenging tasks to address like the efficient comparison of embeddings and their manipulation. In this work, we will discuss about various word and sentence embeddings algorithms, we will select a sentence embedding algorithm, BERT, as our algorithm of choice and we will evaluate the performance of two vector comparison approaches, FAISS and Elasticsearch, in the specific problem of sentence embeddings. According to the results, FAISS outperforms Elasticsearch when used in a centralized environment with only one node, especially when big datasets are included.