Week 6 was relatively uneventful, as I’m still in the very early phases of my code. I’ve been looking into how I should utilize the hypernyms and hyponyms in my code, and this will probably be the biggest issue I face with this method. After all, how far back should I look with the hypernym-hyponym relationships? If one looks far enough back, most words, even words used metaphorically, will end up sharing some sort of hypernym. For example, the phrase all the world’s a stage is metaphorical because world and stage don’t share an immediate hypernym-hyponym relationship. However, venture far back enough, and it can be seen that both words share a common hypernym of physical entity. There needs to be a limit as to how many hypernyms I need to look through in order to return a correct response of metaphorical or literal, and that’s what I’m currently looking into.
In addition, I was actually able to find that the hypernym-hyponym method actually works on adjective-noun phrases, so I will most likely be coding this method to work with AN phrases instead of SVO phrases. This allows me to use the same AN dataset for both the hypernym-hyponym method and the word concreteness method. As I’ve done my research, I’ve found that my initial goal may have been a tad bit ambitious, and I’m starting to cut my work into doable pieces wherever possible–the need for only one dataset would allow me to spend more time coding the actual algorithm and less time having to construct a large SVO dataset.