Movie Scheduling Strategies Using LDA Analysis for Success

Ayan Saraf, Bo Suk Yoon, Lazuli Abel, Peter Laughlin MSBA 46894 M3 - Team G Case Assignment 2 Movie Scheduling & LDA MSBA 46894 M3 - Team G Ayan Saraf: ayananas Bo Suk Yoon: bosuky Lazuli Abel: laabel Peter Laughlin: pmlaughl 1

Ayan Saraf, Bo Suk Yoon, Lazuli Abel, Peter Laughlin MSBA 46894 M3 - Team G Studies have shown that Movie Studios are driven by a mix of intuition and data to back when the right time to release the movie is. The timing is influenced by factors of seasonality, competition and time of year. We use LDA to extract topics and measure similarity between movies, aiding in the strategic scheduling of "The Maze Runner." Movie Topic Clustering 1 The ten topics resulting from an LDA analysis of the movies dataset can be used to group similar movies based on the groups of words viewers have used to describe them. Each of the ten topics, unlabeled clusters resulting from the model, is listed in Figure A across the columns, and the ten movies with the highest similarity scores to each of the topics populate the ten rows of the table. Based on a review of this output, we can arrive at intuitive descriptions of what these topics might represent. See Figure A for those descriptions. Euclidean Distance between Similar Movies 2 To quantify the similarity between "The Maze Runner" and other movies, we employ a Euclidean distance metric in the 10-dimensional space defined by the identified topics. The measure of similarity is derived from the sum of squared differences in topic scores, providing a understanding of the closeness between movies. Figure B outlines the ten movies with the highest similarity scores to "The Maze Runner," presenting a practical application of the Euclidean distance formula. Movie Release Strategy 3 With the insights gained from the LDA analysis and similarity metrics, we extend our analysis to guide the release strategy for "The Maze Runner" in 2014. By considering weekly launch dates, average similarity metrics, and potential competition from preceding weeks, we aim to pinpoint 3 Question 3 2 Question 2 1 Question 1 2

Ayan Saraf, Bo Suk Yoon, Lazuli Abel, Peter Laughlin MSBA 46894 M3 - Team G the most favorable release dates that minimize similarities with other films, thus maximizing the movie's market performance. Based on the clustering analysis conducted on movie keywords, “The Maze Runner” was assigned to cluster 3 along with 51 other movies. To visualize the release pattern of movies in this cluster, a bin diagram(Figure D) was created, showing the trend by release week (week 1 representing the first week of January ). The graph clearly highlights a trend where the majority of movies within this cluster were released during the month of May. Recommended Release Timing 4 Based on weekly similarity scores and number of movies planned to be released during a particular week (for this analysis we consider the period of one week to be the Thursday, the day that typically opens a weekend of releases, through the following Wednesday), there are several weekends that present good opportunities for “The Maze Runner.” However, we cannot ignore the natural seasonality and holiday impact on movie releases. Why release in the first week of November, for example, when movie demand is predictably pent up until later in the month during school breaks (in the US). We also cannot ignore that the distributor of “The Maze Runner,” 20th Century Fox, has many other movies planned for release in 2014 (see Figure E). Depending on the company’s sentiment about the quality of investment of each of these movies, others could certainly get priority release dates in terms of seasonality (this is assuming again that the production company would not want to release one of their own movies in one week). One promising option seems to be the period in mid-May (Week of 05/11) . This is a period of time with relatively few planned releases, all of which appear to be dissimilar to the scores of “The Maze Runner” across the topics derived from our LDA model. This option presumes, however, that all other 20th Century Fox releases would take priority over “The Maze Runner” for seasonality. Another option might be early September (Week of 09/07) for similar reasoning; however, we would recommend a May release over September due to the model’s 4 Question 4 3

Your preview ends here

Eager to read complete document? Join bartleby learn and gain access to the full version

Access to all documents
Unlimited textbook solutions
24/7 expert homework help

Ayan Saraf, Bo Suk Yoon, Lazuli Abel, Peter Laughlin MSBA 46894 M3 - Team G similarity scoring indicating movies more closely related to “The Maze Runner” are being released in September. Another possibility for release is late-march (Week of 03/23) that gives a good balance between releasing movies that have limited overlap with the genre and limited number of total movies releasing that week. We also consider the weeks after the movie release to ensure that “The Maze Runner” gets enough run-time before another blockbuster release to reduce cuts from its profits. Finally, the option of releasing a movie in the holiday season (Week of 12/21) is always present due to it being off work season and a time where audiences collect to the movie theathres. These weeks are highlgited in Figure C that shows the minimum distance and the number of movies being released. Optimize the LDA Model 5 To optimize the release date for "The Maze Runner," we delved into the robustness of the LDA technique by altering the number of topics from the initial 10 to 15 and 20. The objective is to determine how these variations impact the interpretation of topics and the suitability of the model for guiding release strategies. ● Consistency: Certain weeks recur across different topic (e.g., 2014-03-23) suggests a robustness in those periods, indicating key moments of releases ● Nuances and Overlap: With more topics, we observe instances where weeks exhibit similar scores (e.g., 2014-05-20 and 2014-12-21). This consideration of whether these subtleties align with practical marketing decision making. The initial 10-topic model remains intuitive for marketing considerations. The recurring importance of certain weeks reinforces their significance, guiding our recommendations for the release strategy of "The Maze Runner." Using LDA and its optimization techniques ensure that the options to release the movie are optimal for the studio to get maximum exposure. 5 Question 5 4

Ayan Saraf, Bo Suk Yoon, Lazuli Abel, Peter Laughlin MSBA 46894 M3 - Team G Appendix Figure A Figure B Figure C 5

Ayan Saraf, Bo Suk Yoon, Lazuli Abel, Peter Laughlin MSBA 46894 M3 - Team G Figure D Figure E (20th Century Fox 2014 releases) Release Date Film Title Genre Distance 1/17/2014 Devil's Due Horror 0.22 2/28/2014 Son of God Drama 0.17 3/7/2014 Mr. Peabody & Sherman Adventure 0.23 4/11/2014 Rio 2 Adventure 0.24 4/25/2014 The Other Woman Comedy 0.25 5/23/2014 X-Men: Days of Future Past Action 0.63 6/6/2014 The Fault in Our Stars Drama 0.27 6/13/2014 How to Train Your Dragon 2 Adventure 0.29 7/11/2014 Dawn of the Planet of the Apes Adventure 0.30 8/13/2014 Let's Be Cops Comedy 0.22 9/19/2014 The Maze Runner Thriller 0.00 10/3/2014 Gone Girl Thriller 0.66 10/17/2014 The Book of Life Adventure 0.25 11/26/2014 Penguins of Madagascar Comedy 0.27 12/5/2014 The Pyramid Horror 0.19 12/12/2014 Exodus: Gods and Kings Drama 0.06 6

Your preview ends here

Eager to read complete document? Join bartleby learn and gain access to the full version

Access to all documents
Unlimited textbook solutions
24/7 expert homework help

Movie Scheduling_ LDA

Related Documents