Here is the function we talk about in "Semiautomatic dictionary-based classification of tweets about environment"
To create the dictionary you can choose betweet 2 ways:
- manual (1)
- automatic (2)
If you want to proceed with (1) you do not need to upload anything: we recoment you to use it if you do not have any existing set of words about the topic you are interested in or if it is the first time you create a dictionary.
On the contrary (2), we recomend you to use the automatic way if you want to update your existing dictionary. In this case, you need to upload a Twitter dataset, or a sample of it (such as the one you find on GitHub files) please call this file "my_data_twitter" Remember to clean the corpus of my_data_tweet and call the variable containing your cleaned tweets "textClean" If you use the automatic way, you can load hashtags and keywords concerning your topic: please name them "hashtags_TOPIC" and "keywords_TOPIC" Gelerally, create similar files to the one we propose here as example. Do not change variables' name.
Now let's think about OSA list. Do you have a list of tweet talking about your topic of interest? If no, you can use our proposed list, in case you are interested about environment, otherwise, go on Twitter and find some official account which post about the topic of interest. If yes plese call it "tweet_from_OSA" Keep in mind! If you need to create your OSA tweet list remember that you need Twitter API!!