This week, I began to code method 2, which looks at word concreteness and determines whether or not a phrase is metaphorical based on this information. I ran into some issues with method 1, because I need to find a database that can analyze word co-occurrence. It turns out that for AN phrases, method 1 doesn’t use a direct hypernym-hyponym relationship between the adjective and the noun. Instead, method 1 takes the noun and sees whether or not the noun and it’s hypernyms/hyponyms occur frequently with other nouns that typically occur with the adjective. Thanks to Mrs. Rene Flood, a language arts teacher at my school, I found the Corpus of Contemporary American English (COCA), which does display a given word in various contexts–I have yet to implement it into my code, however.
Method 2 is proving to be much easier than method 1–I was able to make XLRD work with Python, a plugin that allows me to read and analyze Excel files within Python. So far, I have two data sheets–one is a collection of around 40,000 commonly used words and their concreteness ratings, and the other is a collection of AN phrases and whether or not they are metaphorical. This is the dataset I will be using throughout my project. I’ve been able to search for and display any given cell within the spreadsheets, and during the coming week, I intend to determine the difference in word concreteness that constitutes a metaphorical phrase. As a reminder, AN phrases with large differences between the concreteness of the adjective and the noun usually make for a metaphorical phrase, while similar concreteness ratings between the two different words usually make for a literal expression.