07-16-2011, 01:20 PM
I have an app idea for you if you haven't already settled on one, though it requires a lot of AI work and it's a long read:
It's a bot that surfs the internet posts messages on forums/facebook, HOWEVER it is not a spambot. It is a oB2Ko Mariobot.
It reads the page's source and parses posts to "read" what people are posting. It puts the strings through some sort of test to find out what the topic is about and what people are saying (somewhat similar to some spambots with the kinda-slightly-vaguely on-topic ads) then makes its own post based on what you would say.
Already hard-coded into the program is a sort of "posting personality", a way of making the Facebook posts or forum posts look like you actually wrote it. A sort of grammar guidelines it has to follow. You would also need a XML file or DB that it reads from to find your interests, like your favorite game or food or vacation spot. It could have several "emotions", where it chooses one based on what keywords were found in the last few posts.
Here's an example:
- Mighty posts that she loves Chinese food.
- [Example Member] posts that they hate Teriyaki Sauce.
- The bot reads the first post
- The bot reads the string "I love Chinese food"
-- The poster is Mighty Jetters, so the "I" is converted to "Mighty Jetters"
-- The bot checks for the verb after the subject "I", which is "love", which the bot associates with a certain positive factor
-- The bot checks for the object after the verb "love", which is "Chinese food". The bot recognizes that the Subject has a positive feeling about the Object
-- The bot reads oB2KoMario.XML and finds out that he has a positive feeling about Chinese food, but not as high.
- The bot reads the second post
- The bot reads the string "I hate Teriyaki Sauce"
-- The poster is [Example Member], so the "I" is converted to "[Example Member]"
-- The bot checks for the verb after the subject "I", which is "hate", which the bot associates with a certain negative factor
-- The bot checks for the object after the verb "hate", which is "Teriyaki Sauce". The bot recognizes that the Subject has a negative feeling about a different Object
-- The bot reads oB2KoMario.XML and finds out that he has a positive feeling about Teriyaki Sauce
- The bot compares the two posts and using Google or something figures out that Teriyaki Sauce is related to Chinese food
- The bot creates a string for the response
-- The bot knows that post 1 is positive and post 2 is negative. The bot is positive about the same thing as post 1 and thus agrees with post 1.
-- The bot starts the string with "Actually, I agree with Might Jetters on this. I like Chinese food". The "actually" comes from identifying the alternate point of view after Mighty Jetters' post
-- The bot is positive about the thing post 2 is negative about and thus disagrees with post 2.
-- The bot adds "and I love Teriyaki Sauce" because it knows there are two objects being talked about.
- The bot posts the string on tSR
obviously it's harder than that but i was just throwing that out there!
My original idea with this was to let the app run in my old age and have it continue posting on forums and FB after I die, heh
It's a bot that surfs the internet posts messages on forums/facebook, HOWEVER it is not a spambot. It is a oB2Ko Mariobot.
It reads the page's source and parses posts to "read" what people are posting. It puts the strings through some sort of test to find out what the topic is about and what people are saying (somewhat similar to some spambots with the kinda-slightly-vaguely on-topic ads) then makes its own post based on what you would say.
Already hard-coded into the program is a sort of "posting personality", a way of making the Facebook posts or forum posts look like you actually wrote it. A sort of grammar guidelines it has to follow. You would also need a XML file or DB that it reads from to find your interests, like your favorite game or food or vacation spot. It could have several "emotions", where it chooses one based on what keywords were found in the last few posts.
Here's an example:
- Mighty posts that she loves Chinese food.
- [Example Member] posts that they hate Teriyaki Sauce.
- The bot reads the first post
- The bot reads the string "I love Chinese food"
-- The poster is Mighty Jetters, so the "I" is converted to "Mighty Jetters"
-- The bot checks for the verb after the subject "I", which is "love", which the bot associates with a certain positive factor
-- The bot checks for the object after the verb "love", which is "Chinese food". The bot recognizes that the Subject has a positive feeling about the Object
-- The bot reads oB2KoMario.XML and finds out that he has a positive feeling about Chinese food, but not as high.
- The bot reads the second post
- The bot reads the string "I hate Teriyaki Sauce"
-- The poster is [Example Member], so the "I" is converted to "[Example Member]"
-- The bot checks for the verb after the subject "I", which is "hate", which the bot associates with a certain negative factor
-- The bot checks for the object after the verb "hate", which is "Teriyaki Sauce". The bot recognizes that the Subject has a negative feeling about a different Object
-- The bot reads oB2KoMario.XML and finds out that he has a positive feeling about Teriyaki Sauce
- The bot compares the two posts and using Google or something figures out that Teriyaki Sauce is related to Chinese food
- The bot creates a string for the response
-- The bot knows that post 1 is positive and post 2 is negative. The bot is positive about the same thing as post 1 and thus agrees with post 1.
-- The bot starts the string with "Actually, I agree with Might Jetters on this. I like Chinese food". The "actually" comes from identifying the alternate point of view after Mighty Jetters' post
-- The bot is positive about the thing post 2 is negative about and thus disagrees with post 2.
-- The bot adds "and I love Teriyaki Sauce" because it knows there are two objects being talked about.
- The bot posts the string on tSR
obviously it's harder than that but i was just throwing that out there!
My original idea with this was to let the app run in my old age and have it continue posting on forums and FB after I die, heh