add idea 1

[Rahmasemma]

No description provided.

[DerekChibbaro]

I really like this idea and could even be used for your own job searching experience. For your visualizations, once you calculate the frequency of key skills or attributes within the category, you could create a word cloud so the audience could view a variety of words and see how often they occur. In terms of potential model ideas, you could see if there is a way to predict the qualifications of certain job postings based on their title or company.

[ernbilen]

EDA-wise this is great but make sure to think about the model too. I think clustering should work, to identify what types of resumes make it to types of companies.

[DerekChibbaro]

This is something that seems to be overlooked at times, and if people are making the effort to report their feedback, they must care about their healthcare experience. I would be curious to see if more reviews are negative since people who had a “normal” or “satisfactory” experience may not feel inclined to give a review. Just a thought if you start getting strange results.

[ernbilen]

• Nice topic—great that you started by explaining the role of a clinical data analyst.
• This role requires both technical and soft skills—good to highlight that.
• Text analysis was great.
• If continuing this project, consider Latent Dirichlet Allocation (LDA) instead of k-means for text clustering.
• Good ethics discussion.
• Good plan for future direction.