I have created a program that puts together (fake) names and words to make... something like gossip.

Some actually made sense and was interesting:

'Ruth is a sniping drugmaker'

'Paul is a exceeding falsifier'

But most didn't:

'Daniel is a sharking burger'

'Bernice is a freshing engender'

'Angela is a disuniting peter'

'Michael is a misapplying harder'

'Donna is a reconfiscating howsoever'

'Irma is a regarding cylinder'

My current algorithm is to randomly select from a list of random words: the third one must end in 'ing', and the last in 'er'.

How can I improve this, so the last word is a good verb, and the second last is a good adjective?

(random names come from http://www.fakenamegenerator.com/order.php)

@Kikinicij (View Post)

 But how can a computer tell verbs and adjectives from other words?

Any sort of generalized structure that applies only to those two?

Also...

'Charles is a stumbling pepper'

@Kikinicij (View Post)

 I use the endings to tell them apart, not because I want those endings...

I will try to look for a list of only verbs, and only adjectives, then.