Abstract | ||
---|---|---|
Learning sentence vectors that generalise well is a challenging task. In this paper we compare three methods of learning phrase embeddings: 1) Using LSTMs, 2) using recursive nets, 3) A variant of the method 2 using the POS information of the phrase. We train our models on dictionary definitions of words to obtain a reverse dictionary application similar to Felix et al. [1]. To see if our embeddings can be transferred to a new task we also train and test on the rotten tomatoes dataset [2]. We train keeping the sentence embeddings fixed as well as with fine tuning. |
Year | Venue | Field |
---|---|---|
2018 | arXiv: Computation and Language | Computer science,Phrase,Natural language processing,Artificial intelligence,Reverse dictionary,Sentence,Recursion |
DocType | Volume | Citations |
Journal | abs/1805.08353 | 0 |
PageRank | References | Authors |
0.34 | 8 | 1 |
Name | Order | Citations | PageRank |
---|---|---|---|
Anson Bastos | 1 | 0 | 1.01 |