Collect data from twitter

Internet activism (also known e-activism or digital campaigning) involves the use of electronic communication technologies such as social media, e-mail, and podcasts for various forms of activism to enable faster and more effective communication by citizen movements. Examples of internet activism include, but are not limited to: LGBT rights, abortion, gun control law, democratic freedoms, climate change, threat of AI, nuclear proliferation, women’s equality, animal and human rights, mental health awareness, political movements and campaigning and access to health etc. 1. Choose one of the internet activism topics listed above and create a list of 5 possible keywords you could use to identify relevant tweets from Twitter. In your answer you SHOULD document any preliminary research you have conducted to select your chosen topic and associated phrases. 2. Using your suggested keywords from part a) and your knowledge of Twitter, collect a series of Tweets surrounding your chosen internet activism topic. You can use either the Twitter streaming or timeline APIs but you MUST justify your choice. Tweets should span a minimum ONE week period. In your answer you MUST provide evidence of how your data was collected (screenshots, code print outs) and describe key methodological steps. 3. Using a suitable example, discuss the role of pre-processing in the context of social media analysis. Identify TWO pre-processing steps relevant to the dataset you created in c) and apply them to your data. In your answer you MUST detail your processing steps and provide evidence that they have been applied. 4. Create a “word cloud” to represent the most commonly used words in the dataset. In your answer you MUST detail your processing steps. 5. Use your processed data file to produce a series of graphs or charts to summarise the following information. I. The number of tweets posted per day II. The number of unique users per day III. The top 10 most active users over the entire period In your answer you MUST detail your processing steps and comment on the results. 6. Using a suitable approach, construct a LDA topic model to identify themes of discussion within your dataset. In your answer you MUST; - Discuss what is meant by topic modelling and explain how your chosen approach works - Provide details of the steps that you have carried out. - Use any tables, graphs and charts you feel are necessary to illustrate your findings - Provide a critical evaluation of your model and discuss one strength and one weakness 7. Choose TWO highly talked about sub-topics within the type of internet activism you have collected tweets for. For example, if you have selected animal rights, one of your sub-topics might be “whale hunting” whilst the other might be the “ivory trade”. Construct a sentiment model for each of your chosen topics and compare and contrast the differences in both polarity and sentiment. In your answer you MUST; - Discuss what is meant by sentiment modelling - Provide details of the steps that you have carried out to build and evaluate your models. - Use any tables, graphs and charts you feel are necessary to illustrate your findings - Provide a critical evaluation of your models and discuss one strength and one weakness