Combining Distributed Vector Representations for Words

Recent interest in distributed vector representations for words has resulted in an increased diversity of approaches, each with strengths and weaknesses. We demonstrate how diverse vector representations may be inexpensively composed into hybrid representations, effectively leveraging strengths of individual components, as evidenced by substantial improvements on a standard word analogy task. We further compare these results over different sizes of training sets and find these advantages are more pronounced when training data is limited. Finally, we explore the relative impacts of the differences in the learning methods themselves and the size of the contexts they access.