Hello,
For the first homework/presentation, do we need to use a dataset of our own or the one provided regarding Wikipedia? (bio-cs-wiki-dataset.csv)
Thank you!
В ответ на Angel Martinez Rodriguez
Re: Topic Modeling Homwork Presentation
Also, regarding the presentations, what is the expected format? Should we go through the notebook presenting our results and explaining roughly the findings and discoveries? Or what is the expected outcome for the project presentations?
Thank you again!
Thank you again!
Hi,
Please consider the provided dataset as a last resort if you are unable to find a more suitable dataset on your own. If you struggle to find one, there are some excellent resources for dataset discovery, most notably:
https://datasetsearch.research.google.com#
https://huggingface.co/datasets
Regarding your second question, we do not have strict requirements for the format of your presentations. However, we want you to "tell a story about your data," meaning you should ideally not only present us with a list of topics but instead show us what topic modeling reveals about the data. This can include (but is not limited to) visualizing topic distributions over documents that might show a rise or decline of topics over time, etc. You can prepare all of this in a notebook and do not need to prepare a slide deck (though that would also be fine).
Best,
Lennart
Please consider the provided dataset as a last resort if you are unable to find a more suitable dataset on your own. If you struggle to find one, there are some excellent resources for dataset discovery, most notably:
https://datasetsearch.research.google.com#
https://huggingface.co/datasets
Regarding your second question, we do not have strict requirements for the format of your presentations. However, we want you to "tell a story about your data," meaning you should ideally not only present us with a list of topics but instead show us what topic modeling reveals about the data. This can include (but is not limited to) visualizing topic distributions over documents that might show a rise or decline of topics over time, etc. You can prepare all of this in a notebook and do not need to prepare a slide deck (though that would also be fine).
Best,
Lennart