A frequent problem in computational linguistics is that passages often use words that the computer simply doesn’t have in its dictionary. Online slang evolves very fast, people use foreign words in English passages, people make typos and invent new abbreviations, etc. You could add new words to the dictionary as fast as you can find them and the next day the program could still be stumped by a new one!
But the program doesn’t have to give up – instead, it can try to work out as much as it can. Various clues can tell a program whether something is a noun or a verb, a person or an inanimate object, etc., and you can even work out more! The following is a webpage where customers have rated their most recent experience at Gelda’s House of Gelbelgarg. Even if you’ve never heard of any of these dishes, you can still figure out some things about them…
A1. Based on the following reviews, attempt to categorize the following items into:
You won’t be able to categorize them with 100% certainty, but use the category that you think is most probable for each. Choose a single category for each word below.
When you are finished with this problem, click on the button below to check your answers.
Checking...
Solution (PDF)
