This is my capstone project during my masters in Smurfit Business School with BearingPoint and Leinster Rugby.

Note: Due to data privacy policy of our client Leinter Rugby club, we have permission to talk about it but not upload images, or videos of our Dashboard and Model.

Deliverables:

The influencer model: Which determined the influencers in the Leinster Rugby community using likes, retweets, and comments.

The interactive social media dashboard: Overall results along with subsections on match day data, sentiments, influencers over time, and tweet searching.

The social media dashboard was the key client deliverable for this project. This dashboard was what Bearing Point will give to Leinster Rugby as a way to strengthen their partnership and generate value for the Leinster Rugby brand. This dashboard will be used by the Leinster Rugby business team in order to better interact with the larger Leinster community. This dashboard was built in Microsoft Power BI and has been linked with not only the data but also the sentiment analysis system and the influencer model.

Power BI Report Pages

First Page: This was the so-called “Landing Page” of the dashboard.

This page provided an overview of all social media categories. It showed numerical values such as the Total Tweets, Likes, Retweets, and Comments.

Underneath those value counts was the influencer graph. Additionally, there was has two word clouds that, using font size, showed the frequency of hashtags and words in a tweet.

Another key piece of this page was the sentiment trend graph, which showed the positive, neutral, and negative trends of all the data, with a date filter to adjust based on user needs. It was also completely interactive. If the user clicked on an influencer account, all graphs would change to stats specific to this account. For instance, if Leinster Rugby was selected in the influencer graph, the sentiment trend graph would change to the sentiment of Leinster Rugby’s specific 19 tweets over time while portions of the sentiment-by-year graph would change to a darker color to illuminate Leinster’s contribution to yearly sentiment. The word clouds and total counts would change to that specific account’s values.

Second Page: This tab described the Twitter activity around match days.

In the top portion of the page, opponent club icons were displayed, and the user was able to select these icons. Other filters included by match result, by type of match, and by location.

In the middle portion of the page, a graph depicting the Twitter sentiment on each match day was present. Finally, the bottom third of the page had the last seven days of sentiment before a match.

This completely interactive dashboard could be used in the following way: a user selected a team from the icon list. The twitter sentiment during match day graph then changed to display only matches dates were Leinster played that specific opponent. Upon selection of a match date in that graph, the bottom graph changed to display the last seven days before that specific match date. This allowed for the complete and thorough analysis of the sentiment surrounding each match.

Third Page This page was specifically designed to further examine the tweets and influencers themselves. The top left graph showed the count and sentiment of tweets overtime. The middle graph, which controlled the other graphics on the page, displayed the average amounts each account gets of likes, comments, and retweets on a post. The top right donut chat showed the percentage of sentiment that was positive, neutral, or negative. The bottom donut chart showed the percentage that likes, comments, and retweets make with concern to engagement. The bottom portion of the graph examined most liked tweets and most re-tweeted tweets. These two graphics provided the account name, the tweet text, and the amount of likes or retweets for each top tweet. This dashboard changed based the account selected in the middle graph. For instance, selecting Leinster Domestic would change all graphs to reflect specific insights into the Leinster Domestic account.

Last Page The last page was the influencer race page. This page was designed to provide a time context to influencer scores. This graphic moved to show the progression of influencer scores over time. This helped provide another way to gain insights from the data.