Python MML Book Recommendation Project

Python
1

Hey all,

I am working on this book recommendation project. I think I almost have it but I am still not getting the correct output. Would anyone mind checking out my Google colab link and giving some feedback on why it isn’t working? Thanks

3

My apologies. I just edited the settings, now you should be able to view. Here’s a new link

4

Your program does work, only not fulfiling all the requirements. Three points can be done:

  1. In cell #4, the two filtering actions need to do at the same time, therefore need to use a single statement( with the two conditions linked by ‘and’) instead of two.

dset = df_ratings[(~df_ratings[‘user’].isin(user_counts[user_counts < 200].index))]
dset = df_ratings[~df_ratings[‘isbn’].isin(isbn_counts[isbn_counts < 100].index)]

  1. As one of the recommended books return by NearestNeighbors is the inputed book itself, if we need 5 recommended books other than the original one, n_neighbors need to be set as 6,

  2. Notice the layouts of the example and testing function, which require the order of the recommended books list to be reversed.

5

If anyone is curious as to how I solved my problems. In cell 4 I was using df_ratings instead of the new dset to get rid of the ISBNs with less than 100 ratings. It had nothing to do with ‘and’ statements.

Also, I made n_neighbors equal to 6 instead of 5 as SzeYeung1 suggested. I then changed the range of my for loop in the get_recommends function to go from 5 to 0 incrementing by -1 and that reversed my list. It works now!

6

This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.