The algorithm sometimes finds very distant relationships to the word you give it. It depends somewhat on the word and whether or not more closely-related words are available. Since "resume" may be seen as both the noun and the verb, related words may be quite distant.

To be quite honest, I was tired of other domain suggestion tools using only the exact word(s) I give, so some degree of distant word relationships is intentional (I considered making it completely random, but I like the idea of choosing a "seed" word as a starting point).