0
$\begingroup$

I'm a working in a small company. The company sells products on a website and they have a python script that runs everyday to attribute a score to each product based on a set of parameters (google analytics events, similar products popularity, price, etc).

The problem is that the scoring outcome is not satisfying, and requiring developers to edit this script arbitrarily, based on business people assumptions, is time consuming and not a proper way to achieve what the business needs.

The ideal solution in my point of view, would be to provide the business team a way to act on the scoring logic in a GUI, integrate what they could come up with, and display the immediate result of applied modifications so they can iterate until they find the right model.

In your experience, what kind of tool/service/library could easily allow non tech users to input formulas on a set of data, which could be synced and integrated in a python script ?

$\endgroup$
0
$\begingroup$

From a Machine Learning perspective, the answer would be to let the non tech users annotate a sample of products with scores and then train a model which predicts scores for all the products. Then the model could be refined and updated by regularly repeating this process, possibly using active learning in order to select the most useful instances and minimize the number of required manual annotations.

Many variants are possible: users could be asked to just say whether they think a score is too high/low, or compare two products and say which one should have a higher score, etc.

$\endgroup$
2
  • $\begingroup$ Thanks for your suggestion, I haven't thought about that, and I love the idea. The process would be indeed easier for business users than coming with their own formula. Do you know any python library that could help to implement such a scoring system ? $\endgroup$ – Matthieu Dsprz Jan 30 at 8:15
  • $\begingroup$ @MatthieuDsprz depends which part you're talking about: one the user side you would need a simple interface to present the user with some products and record the scores they assign. I don't know any particular library but there are probably some, it's very similar to a survey system. On the ML part there are many regression methods and libraries, I would start with simple ones like Decision Tree regression. $\endgroup$ – Erwan Jan 30 at 15:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.