Python Social Media Analytics

Extracting verbatims for keywords

We define a simple function that helps us to visualize verbatims associated with a keyword. The first argument df requires a data frame containing a column 'message', the second one the number of verbatims to show, and the third one is a keyword:

def print_verbatims(df,nb_verbatim,keyword): 
 
    verbatims = df[df['message'].str.contains(keyword)] 
     
    for i,text in verbatims.head(nb_verbatim).iterrows(): 
        print(text['message']) 

We can do simple searches by using the following syntax:

print_verbatims(df_comments,5,'google') 

It is worth noting that some short keywords might be a part of a longer string, so in order to get precise tokens, we use a regex word boundaries \b: print_verbatims(df_comments,5,'\\bgoogle\\b') ...