All Data
After having a look through the data as a whole as well as each of the four smaller “communities” there are a few key points that I found interesting about the song selection data. When looking at the four different communities, I was mostly interested in looking for the outliers- the songs that were only chosen by one person in the group. Each community features members who had all of their songs selected by at least one other person, as well as at least one member who had at least one outlying song that did not connect to any other group members. However, looking at the data as a whole which included all four communities, I was pleasantly surprised to see that there is not a single outlying song.
I wanted to see which, if any, of the 27 songs on the master list had not been chosen by anyone at all. I used the search function (command + F) on my computer to make sure I was not missing anything within the web of data and I searched each song and was very surprised to find that all 27 songs were present in the data AND, as previously stated, all were chosen by at least two people as there are no outliers.
The three songs that were seemingly the least popular choices which were only chosen twice are Track 8: Men’s House Song, Track 17: Well-Tempered Cavalier, and Track 27: String Quartet No. 13 in B Flat. I, like most people, did not include any of these songs on my list. My reason was that I was looking for only one song from each continent and there was another song I liked better from these songs’ continent of origin, but I am curious to know what my classmate’s reasons are. Did these songs simply fall through the cracks or are they more or less disliked?
Community 3
One thing that stood out to me when looking at each community separately was how Selene did not seem to fit in where she was placed in community 3. When looking at only community 3, Selene has 9 songs that are outliers that do not connect with anyone else, which makes me wonder why she was placed in that group. When you look at the data all together, Selene is right in the centre showing that she had very common selections to a lot of our classmates.
Hi Erin,
Your observation about Selene is very intriguing. I also noticed that when a person is placed in a community, it does not necessarily mean that he/she has the most connections with everyone in the community. I’m in community 3 with Yijun and Junel, sharing 5 and 4 vertices with them respectively. I share 5 songs with you, but we are not placed in the same community.
I’m guessing the software does not “optimize” for each person. Instead, it counts all the vertices produced by our class and makes selections based on everyone’s choices. Any other guesses? I wish Palladio can share its algorithms, so we know exactly why a person is placed in a certain community.