DISSERTATION PROJECT

Social media has become a major part of society, with many depending on it for their news, communication with others and source of entertainment. Due to the quickness of social media’s rise, there is a number of impacts it is having on society that are not being address fast enough or at all. The purpose of the report is to explore this topic of how social media data and algorithms are impacting society and find potential solutions using data from social media. Natural language processing implementations can be used in multiple ways to limit the negative impacts and promote the positives. To do so different types of NLP must be considered. Sentiment analysis can be used to provide an opinion on the data shared on social media, especially to control the role of emotional contagion plays in people’s mental health as a direct result of social media. To do so, an application is developed to find the current sentiment score of a user’s latest tweets, to determine what the user’s current mindset maybe, which can then be used to influence that user’s algorithm. Another major impact is the spreading of false information as there is currently no implemented checks on content data containing potential false information, as it first needs to be reported by a user. By implementing a form NLP to analyse data before it is upload, so that it could be marked as potentially misleading, would have a massive impact as false information tends to spread quickly on social media. The two applications created and used in this report are done using Python. There is a major consensus that social media data and algorithms are having an impact on society, however determining if this is positive or negative is important and how to limit the negative impact can be done using natural language processing implementations.