Chapter 7 Chapter 7 Conclusion

Our analysis and detailed exploration of the IMDb dataset has allowed us to obtain very interesting insights to answer our three objective questions.

7.1 Insights

7.1.1 Objective Question 1

For the objective question 1, we derive the following basics conclusions:
1. With the progress of time, the number of adult titles that were produced by the director started increasing, Though, this trend was not of a continuous increasing one but it is observed that as compared to very early years, the number of adult title types releasing increased. Thus, we can conclude that we progressing time and generations, people started likening media content that had some adult content and the society was more open to such entertainment content as compared to the old times.
2. The surprising insight that we got from this question is that the proportion of the number of votes received by non adult title types were still very large as compared to the proportion of number of votes received by the adult type in any given year in the dataset. Although, more adult title types started releasing over time but it is concluded that people were more tend to give votes for movies, TV series, videos, etc that were non adult in nature rather than voting for the adult media types.

7.1.2 Objective Question 2

The largest number of observations come from the movie and tv title types. This may be because there is a larger market for TV series and movies due to streaming services like Netflix and Hulu. Many of these works are therefore marketed at a global scale. As stated before, we saw that there are more votes for tv related title types rather than movies. This may be due to the considerable time that an audience member would invest to understand and watch a series or an episode. There is more room for the audience member to grow attached to the characters and see their development. This also leads to higher quality works in the market. Short films on the other hand may be more local with less well known artists. may have trouble getting publicity. Game shows are less popular amongst the youngest generation which may cause them to lose a large number of votes. Video game types are most likely only watched by those who play video games rather than the general public.

7.1.3 Objective Question 3

For the titleTypes of tvMiniSeries, tvSeries, and tvSpecial, there is only 1 pair of director-writer who have an average rating of 10 and have worked together for more than 5 times. The titleTypes short, tvEpisodes, and video have many director-writer pairs that have average rating 10 and have worked for more than 5 times together. Surprisingly, the plot reveals that there is no pair of director-writer that have worked together for more than 5 years and achieved an overall average rating of 10 for the titleType movies, this suggests that there could be two scenarios for this: first there would be pairs that have average rating of 10 but the pair might have worked for less than 5 times or second there could be no movies that have average ratings of 10.

7.2 Limitations

Our largest setback was in dealing with the large scale of the data. We wanted to look at different directors and analyze more trends in what/who seems to bring in the highest votes and ratings. However, the large amount of directors slowed down our work immensely and made it difficult for us to plot in a readable fashion.

7.3 Future Directions

• The IMDb dataset has a names dataset that contains the first, second, and tertiary professions or the individuals listed in the database. It would interesting to look deeper into how the professions of an individual can change their impact on a work, i.e., comparing the average ratings of individuals who have experience as directors and writers vs directors and actors.
• Another dataset in the database had spatial data that would have been useful in seeing trends in how different title types are received based on location.