The Roaming Abstract bot uses three source "grammars" to create tweets. Source grammars are the equations that Tracery uses to plug in different words and phrases to create unique sentences. The three grammars that I am using are:
"[Blank] in [blank] in a [blank]." - more specifically - "[Object] in [color] in a [type of room]."
"[Blank], [blank], [blank]." - more specifically - "[Type of industry], [noun], [noun]."
"[Blank] meets [blank] in the style of [blank]." - more specifically - "[Noun] meets [noun] in the style of [specific art movement]."
These three basic formulas give the bot a rigid feel - it only posts from one of the three grammars, at an interval of 10 minutes apart. The lists of objects, nouns, colors, industries, rooms, and art movements were found, pre-written, on this site: https://github.com/dariusk/corpora/tree/master/data
For this project, the rigidity will help me compartmentalize the different types of responses I want to create. The "[Object] in [color] in a [type of room]." and "[Noun] meets [noun] in the style of [specific art movement]." grammars lend themselves more towards visual artistic responses, such as paintings and drawings, assemblages and sculptures. The "[Type of industry], [noun], [noun]." grammar seems to lend itself to a more abstract type of response. For these types of tweets I will be responding sonically. I plan to post break-downs of my sonic creations along side of each tweet from this category.
"[Blank] in [blank] in a [blank]." - more specifically - "[Object] in [color] in a [type of room]."
"[Blank], [blank], [blank]." - more specifically - "[Type of industry], [noun], [noun]."
"[Blank] meets [blank] in the style of [blank]." - more specifically - "[Noun] meets [noun] in the style of [specific art movement]."
These three basic formulas give the bot a rigid feel - it only posts from one of the three grammars, at an interval of 10 minutes apart. The lists of objects, nouns, colors, industries, rooms, and art movements were found, pre-written, on this site: https://github.com/dariusk/corpora/tree/master/data
For this project, the rigidity will help me compartmentalize the different types of responses I want to create. The "[Object] in [color] in a [type of room]." and "[Noun] meets [noun] in the style of [specific art movement]." grammars lend themselves more towards visual artistic responses, such as paintings and drawings, assemblages and sculptures. The "[Type of industry], [noun], [noun]." grammar seems to lend itself to a more abstract type of response. For these types of tweets I will be responding sonically. I plan to post break-downs of my sonic creations along side of each tweet from this category.
Comments
Post a Comment