Added a section for Multimodal Metaphor Recognition

Added a section for some of the research on Multimodal (specifically visual and text) Metaphor Recognition.

multimodal.md

@@ -31,4 +31,16 @@ The MOSI dataset ([Zadeh et al., 2016](https://arxiv.org/pdf/1606.06259.pdf)) is
 | bc-LSTM (Poria et al., 2017) | 80.3%  | [Context-Dependent Sentiment Analysis in User-Generated Videos](http://sentic.net/context-dependent-sentiment-analysis-in-user-generated-videos.pdf) |
 | MARN (Zadeh et al., 2018) | 77.1%  | [Multi-attention Recurrent Network for Human Communication Comprehension](https://arxiv.org/pdf/1802.00923.pdf) |
 
-[Go back to the README](README.md)
+## Multimodal Metaphor Recognition
+
+[Mohammad et. al, 2016](http://www.aclweb.org/anthology/S16-2003) created a dataset of verb-noun pairs from WordNet that had multiple senses. They annoted these pairs for metaphoricity (metaphor or not a metaphor).
+
+[Tsvetkov  et. al, 2014](http://www.aclweb.org/anthology/P14-1024) created a dataset of adjective-noun pairs that they then annotated for metaphoricity.
+
+Both datasets are in English.
+
+| Model                                                        |    Score     | Paper / Source                                               | Code        |
+| ------------------------------------------------------------ | :----------: | ------------------------------------------------------------ | ----------- |
+| 5-layer convolutional network (Krizhevsky et al., 2012), Word2Vec | 75.0%, 79.0% | [Shutova et. al, 2016](http://www.aclweb.org/anthology/N16-1020) | Unavailable |
+
+[Go back to the README](README.md)