Upgrade duplicate finding logic with edit distance

[vinayak-mehta]

We could have a threshold (0.5 or 0.7) above which a conference could be flagged as a duplicate during import.

[Sangarshanan]

We could use Levenshtein or difflib.SequenceMatcher. There is a stackoverflow answer comparing the Performance and Tradeoffs of both

[nishant-sethi]

@vinayak-mehta is this issue still open?

[vinayak-mehta]

@Sangarshanan was working on this. If he is not working on it, then you can take this up :)

[nishant-sethi]

@Sangarshanan can you please confirm if you are working or not.

[Sangarshanan]

I am not working on this currently. You could take this up if you want @nishant-sethi

[nishant-sethi]

@vinayak-mehta I'll pick it up. @Sangarshanan can you please tell me in which file changes are required.
@vinayak-mehta Can you please me the overview about what change is required.

[Sangarshanan]

Thank you for picking this up @nishant-sethi :)

You could add it as a utils functions that could get picked up by import cli command. Also I would suggest you use Levenshtein similarity and set the threshold a bit higher, around 0.9

[nishant-sethi]

@Sangarshanan Thanks for the suggestion. I'll try to work on this today and raise PR based on which you can provide feedback.

[JosemyDuarte]

Hi pps! Do you keep working on this @nishant-sethi? I can take it if it's free.

[nishant-sethi]

@JosemyDuarte you can take this up

[JosemyDuarte]

Hi @vinayak-mehta! I just opened a PR for this. Whenever you can, please let me know your feedback.