2/27/19 - Githubissues

bellecarrell / twitter_brand

In developing a brand on Twitter (and social media in general), how does what you say and how you say it correspond to positive results (more followers, for example)?

0 stars 1 forks source link

2/27/19 #99

Closed bellecarrell closed 4 years ago

bellecarrell commented 5 years ago

move things around to address confusion around purpose of paper
study for sentiment
define strategy post more (# times posted/day, etc)
add that dataset will be available
take words, generate tweets, run through model for prediction of success

bellecarrell commented 5 years ago

@abenton

bellecarrell commented 5 years ago

add appendix to paper
ablation studies for different strategies (lexical n-grams, studying the higher-level strategies, such as the posting frequency, the type of posts, the length of posts, the sentiment of the posts) [test explanatory power]
generate tweets + predict popularity

bellecarrell commented 5 years ago

Dirk:

refocus goal of paper: compare strategies, point out value for bloggers, make data available
data will be made available
other social media source? difficult...
predictive vs explanatory: generate tweets with the successful words contained, test their effect?
explain "strategy" and its components, and explain how we test it
explain annotation scores
improve domain prediction
add back appendix
explain data statement (cite!)

Addtl. Exps:

ablation studies (n-grams vs frequency vs time vs sentiment)
measure sentiment
measure human performance in popularity prediction

bellecarrell commented 5 years ago

next step: run ablation studies