Ask, and you shall get. Scrapes Twitter/Reddit queries to give you only the highest quality answers.
Run the Makefile by calling make, followed by make ask. Respond with whether or not you would like relationship advice (highly recommended!), and if you'd like a serious response (prioritize upvoted answers or downvoted answers?). For Obama speech generation, run make obama.
-
Run the python file as follows, where the weights flag tells the program to output each comment's weight at the top of its file.
python answer_question.py --weights -
When the program prints out a message, respond with whether or not you would like relationship advice and if you'd like a serious response. Then, input your question and hit enter.
- Run "java Markov" after compiling. Input a -h flag to see the help message, which is also outputted here for your convenience:
Usage: Markov ...
(optional) \[-w or -weight\] \[read_mode or DOUBLE_NUMBER\]
(required) \[-f or -file\] \[FILE_NAME or DIR_NAME\]
(recommended) \[-n or -num_words\] \[INTEGER_NUMBER\]
(do NOT use with -n!) \[-t or -twitter\]
- run
python tweet_collector.py. If you know the Twitter username of a person, runpython tweet-collector.py -u $(username)instead (replace$(username)). - After the tweets has been stored in a file format $(username)_tweets, feed that file into Markov Chain by running
java Markov -f $(username)_tweet -t.-tflag will output a Tweet-like message into the console.
NOTE: The program may run into issues with sending too many requests to the server. If this occurs, just wait a little while and execute the program again with the same query string.