Incorporating Both Distributional and Relational Semantics in Word Representations
This work addresses the challenge of creating more comprehensive word embeddings for natural language processing applications, though it appears incremental as it builds on existing methods.
The paper tackled the problem of enhancing word representations by integrating both distributional and relational semantics, using ADMM to optimize objectives on raw text and WordNet, resulting in improvements in some cases on tasks like knowledge base completion and analogy tests.
We investigate the hypothesis that word representations ought to incorporate both distributional and relational semantics. To this end, we employ the Alternating Direction Method of Multipliers (ADMM), which flexibly optimizes a distributional objective on raw text and a relational objective on WordNet. Preliminary results on knowledge base completion, analogy tests, and parsing show that word representations trained on both objectives can give improvements in some cases.