Using R, a software environment for statistical computing, it's possible to extract data from social network like Twitter or Facebook, thanks to the application of recently developed packages. After the data extraction, that mostly concerns the Italian banks listed on the Stock Exchange, it will follow the "Sentiment Analysis", valuating the expressed judgments and giving them a positive, negative or neutral value, depending on the opinion. It's then possible, thanks to the R codes, to evaluate the whole extracted samples in order to observe what people think and say on social network. Finally, the evaluation obtained will be compared with the credit risk analysis traditionally made by the rating analysts on the interested subject (in our case, the Italian banks).
Grazie all'utilizzo di R, un software molto utilizzato in ambito statistico, è possibile estrarre dati da social network come Twitter o Facebook grazie all'applicazione di pacchetti recentemente sviluppati. All'estrazione dei dati, che riguarda in particolare le banche italiane quotate in borsa, seguirà una fase di "Sentiment Analysis" dove si valutano i giudizi espressi assegnando un valore positivo, negativo o neutro in base al carattere dell'opinione. E' poi possibile, grazie all'utilizzo di codici R, valutare tutti i campioni estratti per osservare cosa emerge dalle opinioni che le persone esprimono sui social network. Infine, la valutazione così emersa viene paragonata con l'analisi di rischio di credito comunemente usata dai rating analysts per il campione di interesse (le banche italiane nel nostro caso).
TWITTER SENTIMENT ANALYSIS TO MEASURE STRATEGIC RISK OF ITALIAN BANKS: THE UNICREDIT CASE STUDY
MAZZARELLA, MARTA
2014/2015
Abstract
Using R, a software environment for statistical computing, it's possible to extract data from social network like Twitter or Facebook, thanks to the application of recently developed packages. After the data extraction, that mostly concerns the Italian banks listed on the Stock Exchange, it will follow the "Sentiment Analysis", valuating the expressed judgments and giving them a positive, negative or neutral value, depending on the opinion. It's then possible, thanks to the R codes, to evaluate the whole extracted samples in order to observe what people think and say on social network. Finally, the evaluation obtained will be compared with the credit risk analysis traditionally made by the rating analysts on the interested subject (in our case, the Italian banks).È consentito all'utente scaricare e condividere i documenti disponibili a testo pieno in UNITESI UNIPV nel rispetto della licenza Creative Commons del tipo CC BY NC ND.
Per maggiori informazioni e per verifiche sull'eventuale disponibilità del file scrivere a: unitesi@unipv.it.
https://hdl.handle.net/20.500.14239/4383