Teens Wander Around a House, by Darius Kazemi

dariusk commented 10 years ago

Hi everyone. I have no clue what I'm going to do yet, but I'm tracking my progress here. Excited for this :smiley:

dariusk commented 10 years ago

Okay, here's my first novel! http://tinysubversions.com/nanogenmo/novel-1.pdf

It uses dreams from http://dreambank.net/ and follows this structure:

Act 1: random dreams of girls age 9-11
Act 2: random dreams of girls age 12-14
Act 3: random dreams of girls/women age 14-20
Interlude: random dreams of a child molester (really), run through a text obfuscator I wrote
Act 4: random dreams of women age 20-29
Coda: markoved text, merging the dreams of girls age 9-11 with the child molester

I'll post the code later.

dariusk commented 10 years ago

Basically the idea here was: could I approximate a sense of narrative forward movement by moving between corpuses of the dreams of girls of ascending age, with a creepy/awful interlude to break things up.

dariusk commented 10 years ago

Okay, I've been playing around with dialogue. Here's an example:

"What is the best lesson you've had using epic citedel?"

Delinda squirmed. "Realizing you can mix science and humanities."

"Why do you think you're so relevant?"

Delinda flinched. "Because i didn't pull all nighters for material not relevant to me in college right now."

"What's the best tweet you read today?" demanded Celinka.

Delinda sighed. "I think my last rt."

Full example here: https://gist.github.com/dariusk/b74ecc5ec6b83d62cdba

It grabs three kinds of question and answer:

subtweets in the form of "why do you / why would you", along with tweets that have "because i"
- I filter this further by grabbing the longest word from the question, and finding a "because i" tweet that contains that word
tweets that ask, "What is the best ___?" It grabs the first word after "best" and finds tweets that say " [whatever] is the best [firstword]", setting "[whatever]" to the answer
same as above, but "your favorite" and "my favorite" as the pair

enkiv2 commented 10 years ago

That is very cool. Do you have some sort of knowledge-base for categorizing what would be an appropriate response? (Something storing categories and their members, like maybe the OpenCog/Cyc/NELL ontologies?)

On Thu, Nov 7, 2013 at 3:46 PM, dariusk notifications@github.com wrote:

Okay, I've been playing around with dialogue. Here's an example:

"What is the best lesson you've had using epic citedel?"

Delinda squirmed. "Realizing you can mix science and humanities."

"Why do you think you're so relevant?"

Delinda flinched. "Because i didn't pull all nighters for material not relevant to me in college right now."

"What's the best tweet you read today?" demanded Celinka.

Delinda sighed. "I think my last rt."

Full example here: https://gist.github.com/dariusk/b74ecc5ec6b83d62cdba

It grabs three kinds of question and answer:

subtweets in the form of "why do you / why would you", along with tweets that have "because i"

I filter this further by grabbing the longest word from the question, and finding a "because i" tweet that contains that word

tweets that ask, "What is the best ___?" It grabs the first word after "best" and finds tweets that say " [whatever] is the best [firstword]", setting "[whatever]" to the answer

same as above, but "your favorite" and "my favorite" as the pair

— Reply to this email directly or view it on GitHubhttps://github.com/dariusk/NaNoGenMo/issues/2#issuecomment-28004745 .