Semantic Regularities in Document Representations
This work addresses the problem of semantic reasoning at the document level for NLP researchers, but it is incremental as it extends known word-level concepts to documents.
The paper investigates whether document representations exhibit linear semantic regularities similar to word embeddings, by designing a document analogy task and evaluating state-of-the-art models. Results show neural embedding-based methods outperform conventional ones on this task, with preliminary explanations provided.
Recent work exhibited that distributed word representations are good at capturing linguistic regularities in language. This allows vector-oriented reasoning based on simple linear algebra between words. Since many different methods have been proposed for learning document representations, it is natural to ask whether there is also linear structure in these learned representations to allow similar reasoning at document level. To answer this question, we design a new document analogy task for testing the semantic regularities in document representations, and conduct empirical evaluations over several state-of-the-art document representation models. The results reveal that neural embedding based document representations work better on this analogy task than conventional methods, and we provide some preliminary explanations over these observations.