Skip to content

Latest commit

 

History

History
75 lines (59 loc) · 5 KB

guestbook_cassie.md

File metadata and controls

75 lines (59 loc) · 5 KB
Raw

Gender Dynamics in Animated Movie Dialogue

Explore my project here!

So far, my project is mostly data organization, and right now my data really resembles a hydra (each time I fix one problem in my data, three more rear their ugly heads).

Any and all feedback is really appreciated!!!!

Ting-Wei

  • Strength: Interesting topic and nice data cleaning. Clear project explanation.
  • Improvement: Shorten the length of jupyter notebook, and keep the main point.
  • Thing I learned: I never think about that we can detect the gender of movie characters by their dialogue. It is really interesting to me. I learned a lot of data cleaning methods from this project.

Cassie's Response:

Thanks for visiting Ting-Wei! I agree, I can definitely shorten up some of my notebooks/clean up their organization! I might split a few up to make them more readable.

Patrick

  • One thing I liked : Wow you watched Moana and typed all the lines in? Nice lol. Overall, I like your plan and your code seems well written.
  • One area of improvement : Why'd you get rid of Lion King? There isn't a princess but there's Nala. Also, imo you may want to consider how significant the messiness of the disney corpus is. Your data will never be perfect, and it may not be worth the time and effort to go through all of the data and fix everything, depending on how messy it is. Just something to think about.
  • One thing I learned : A method to create a dataframe of lines of speech from a movie script.

Cassie's Response:

Thanks for visit, Patrick! I took your advice to heart when prepping my second project report. I can't fix ALL the data, so I ended up only fixing issues as I came across them during my annotation processes.

Katie

  • One thing I liked : Such a cool topic! I also like how you really walk through the code so we can understand your process and what's going on. I'm also impressed by how much work goes into the cleaning of the data.
  • One area of improvement : Will you be doing any machine learning? I understand why you wouldn't, since your task is already so huge! But if you do, an easy(/ier) one would be identifying the gender of the speaker.
  • One thing I learned : The use of the "partial" function was interesting! A cool way to reuse code.

Cassie's Response:

Thanks for the visit, Katie! I like the idea of using machine learning to identify gender. If I have time to do this after my other work, I will definitely try it out!

Matt

  • One thing I liked : Interesting idea. I enoyed Disney, Dreamworks and other related animated movies when I was young, so I can appreciate the content being used. Also, I like the large amount of content being processed as well.
  • One area of improvement : Relative file paths should be used instead of the file paths you have now. Gives your filee save/load commands a cleaner look, and makes it so that you can run your code can be run on different machines with little hassle.
  • One thing I learned : Streamlining multiple different corpora using slightly different methods for each. Also a nice refresher on how to use the re module in python.

Cassie's Response:

Thanks for visiting my project, Matt! I've always been a huge fan of animated movies, so this corpus makes a sometimes frustrating project a little more fun. Also, you're right about the relative file paths. I always just copy the path from my directory, and rarely bother to edit it in my code. Thanks for the heads up!

John

  • One thing I liked: Your project idea is so. COOL! Also, you are approaching this monster of a task very methodically and logically, so it's super easy to see what you're going for (especially with your regexp). Keep up the great work!
  • One thing that could be improved: I know you're not finished yet, but it would be interesting to see how the data is currently conforming/non-conforming to your hypotheses. It's always nice to recenter and make sure that you're not missing the bigger picture.
  • One thing I learned: .sub seems pretty interesting! I'm going to need to check that out in the future.

Cassie's Response:

Thanks for visiting John! And thanks for the advice! I do tend to get lost in data clean-up...

Eva

  • What was done well: Good explanation and organization throughout your project! The project plan and progress report gave me a good idea of your project goals and methods. Really great job cleaning up and annotating your data. Data visualizations are great too.
  • Improvements and suggestions: Some of the links in your progress report don't work - looks like you moved all your Disney and Dreamworks code into separate folders? Because of this, I wasn't totally sure what Jupyter notebooks to look at first, though I was still able to skim over all of them and understand your process.
  • What I learned: I would have expected Disney movies to be worse at including female characters than Dreamworks movies, but it looks like Dreamworks movies are more male-dominated.