Social Networks and, in particular Twitter, have been proved many times to be a useful source of information for both scholars and industries. In particular, industries are typically interested in feedbacks about the impact of social marketing campaigns on their users. Moreover, for many businesses Twitter can also be a useful companion platform that helps to detect service outages or issues by studying the interaction of the users with the official Twitter page of the firm. However, to fully exploit the advantages brought about this powerful interaction and sharing platform, it is important to understand which are the strategies typically adopted by industries to promote their products and acquire user consensus. To provide a contribution in this setting, this thesis focuses on four Broadcast Streaming Providers, namely Netflix US, Netflix IT, NowTV UK, and NowTV IT, and it describes a comparative analysis performed to understand the dynamics in the interaction with users and the impact of different interaction strategies to the achieved user engagement. Starting from previous studies on the life-cycle of technology adoption, the four businesses under investigation have been chosen to have different technology adoption levels as well as the possibility of studying the diverse strategies of the same service in the headquarter country (US for Netflix and UK for NowTV) and the Italian counterpart. This thesis will explore different machine learning techniques, including NLP and Text Mining approaches, to compare the behavior of the four Twitter pages, namely @NowTv, @NowTv_it, @Netflix, and @Netflix_it corresponding to businesses under analysis. The results reported in this thesis represent a premise to understand the success strategies that can be adopted by Broadcast Service Providers to promote their products and services using their social network accounts. The last part of this thesis presents an advanced data analytic tool designed to monitor the interaction of users with a target Twitter page. This tool has been designed by leveraging solutions typical of Big Data scenarios and it allows for the processing of streaming data directly received from Twitter to feed a monitoring custom dashboard.
Utilizzo dei dati di Twitter per comprendere l'evoluzione dei servizi di trasmissione in streaming: un'analisi di Netflix e NowTV. I social network e, in particolare, Twitter, si sono dimostrati molte volte un'utile fonte di informazioni sia per gli studiosi che per le industrie. In particolare, le industrie sono tipicamente interessate ai feedback sull'impatto delle campagne di social marketing sui propri utenti. Inoltre, per molte aziende Twitter può anche essere un'utile piattaforma che aiuta a rilevare interruzioni o problemi del servizio studiando l'interazione degli utenti con la pagina Twitter ufficiale dell'azienda. Tuttavia, per sfruttare appieno i vantaggi apportati da questa potente piattaforma di interazione e condivisione, è importante capire quali sono le strategie tipicamente adottate dalle industrie per promuovere i propri prodotti e acquisire il consenso degli utenti. Per fornire un contributo in questo contesto, questa tesi si concentra su quattro Broadcast Streaming Provider, ovvero Netflix US, Netflix IT, NowTV UK e NowTV IT, e descrive un'analisi comparativa eseguita per comprendere le dinamiche nell'interazione con gli utenti e l 'impatto di diverse strategie di interazione per il coinvolgimento dell'utente raggiunto. Partendo da precedenti studi sul ciclo di vita dell'adozione della tecnologia, le quattro aziende oggetto di indagine sono state scelte per avere diversi livelli di adozione della tecnologia nonché la possibilità di studiare le diverse strategie dello stesso servizio nel paese sede (USA per Netflix e UK per NowTV) e la controparte italiana. Questa tesi esplorerà diverse tecniche di machine learning, inclusi gli approcci di NLP e Text Mining, per il comportamento delle quattro pagine Twitter, ovvero @NowTv, @NowTv_it, @Netflix e @Netflix_it corrispondenti alle aziende in analisi. I risultati riportati in questa tesi rappresentano una premessa per comprendere le strategie di successo che possono essere adottate dai Broadcast Service Provider per promuovere i propri prodotti e servizi utilizzando i propri account di social network. L'ultima parte di questa tesi presenta uno strumento avanzato di analisi dei dati progettati per monitorare l'interazione degli utenti con una pagina Twitter di destinazione. Questo strumento è stato progettato facendo leva su soluzioni degli scenari Big Data e permette di elaborare i dati in streaming ricevuti direttamente da Twitter per popolare una dashboard di monitoraggio personalizzata.
Using Twitter data to understand the evolution of Broadcast Streaming Services: A Netflix and NowTV analysis.
ARAZZI, MARCO
2019/2020
Abstract
Social Networks and, in particular Twitter, have been proved many times to be a useful source of information for both scholars and industries. In particular, industries are typically interested in feedbacks about the impact of social marketing campaigns on their users. Moreover, for many businesses Twitter can also be a useful companion platform that helps to detect service outages or issues by studying the interaction of the users with the official Twitter page of the firm. However, to fully exploit the advantages brought about this powerful interaction and sharing platform, it is important to understand which are the strategies typically adopted by industries to promote their products and acquire user consensus. To provide a contribution in this setting, this thesis focuses on four Broadcast Streaming Providers, namely Netflix US, Netflix IT, NowTV UK, and NowTV IT, and it describes a comparative analysis performed to understand the dynamics in the interaction with users and the impact of different interaction strategies to the achieved user engagement. Starting from previous studies on the life-cycle of technology adoption, the four businesses under investigation have been chosen to have different technology adoption levels as well as the possibility of studying the diverse strategies of the same service in the headquarter country (US for Netflix and UK for NowTV) and the Italian counterpart. This thesis will explore different machine learning techniques, including NLP and Text Mining approaches, to compare the behavior of the four Twitter pages, namely @NowTv, @NowTv_it, @Netflix, and @Netflix_it corresponding to businesses under analysis. The results reported in this thesis represent a premise to understand the success strategies that can be adopted by Broadcast Service Providers to promote their products and services using their social network accounts. The last part of this thesis presents an advanced data analytic tool designed to monitor the interaction of users with a target Twitter page. This tool has been designed by leveraging solutions typical of Big Data scenarios and it allows for the processing of streaming data directly received from Twitter to feed a monitoring custom dashboard.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/20.500.14239/12185