Fix/real keyword extraction analytics

[OmanshiRaj]

CLOSES #49

Summary

Fixes inaccurate analytics where the "Common Keywords from Comments" section always displayed the static keywords fast, examples, unclear, and good regardless of actual student feedback.

Changes Made

• Removed hardcoded placeholder keywords.
• Added keyword extraction utility based on comment text.
• Normalized comments by converting text to lowercase and removing punctuation.
• Filtered common English stop words.
• Counted word frequencies across all comments.
• Returned the top N most frequent keywords.
• Added handling for empty or missing comments.

Before

Student comments such as:

• "too slow"
• "loved the diagrams"
• "great explanation"

would always display:

• fast
• examples
• unclear
• good

After

Keywords now reflect actual student feedback, for example:

• slow
• diagrams
• explanation
• loved

Testing

• Submitted feedback containing custom comments.
• Verified that analytics displays words present in the submitted comments.
• Verified stop words are excluded.
• Verified empty comment datasets return an empty keyword list.

Fixes #{issue_number}

[vercel]

@OmanshiRaj is attempting to deploy a commit to the Rishima's projects Team on Vercel.

A member of the Team first needs to authorize it.

[rishima17]

@OmanshiRaj pls add relevant screenshots also.