2
$\begingroup$

I'm trying to explore an use-case in ML but stuck at a point. May i please request your advise please.

Have a service desk web application for logging tickets, which is essentially a form having various fields like -

subject,
content, 
username, 
emailid,
category - {hardware, application, datafix, mobileapp, etc.}, 
service group - {AAA, BBB, CCC, DDD, etc.}, 
domain - {email, walkin, phonecall, etc.},
priority - {high, medium, low}
#(Apologies for the poor quality of sample data provided above.)

Based on this info, the ticket is then manually assigned to respective team owners for resolution.

My intent is to use ML - based on the above fields, predict the Team who will work on this ticket. (Team ex. HR or IT or Desktop support or Pantry or Facilities, etc. )

  1. Can this use-case be categorized as Multi-class classification problem?
  2. The field values are stored as words in database. How can it be fed to my ML as numbers?

Please advise.

$\endgroup$
0
$\begingroup$

Welcome to the DataScience Stack exchange.

Since your target variable has many categories your problem is Multi class classification. On how to implement and find a solution for this problem you can find on google. The following 2 links are for reference:

  1. Link-1
  2. Link-2

Here I'm not giving much information on which algorithm to use. If you need any additional help you can raise one more question so that we can help you better as we don't now much about your data and business problem.

So, to answer your question 2, you can convert them into number it is one line code,I've added the link for your reference.

Please let us know if you need any additional information.

$\endgroup$
  • 1
    $\begingroup$ Thanks for your response. Much appreciated. I'll go through it and revert if any doubts. I've done some practice hands-on with ML+Py, now undergoing a real use-case. :) $\endgroup$ – ranit.b Jan 14 at 14:20
  • $\begingroup$ Your link-2 was really useful and exactly mentions what I was looking for. Many thanks. It covers aspects of - Text representation (using Bag-of-words and TFIDF), and n-grams to understand the correlation among the words. $\endgroup$ – ranit.b Jan 15 at 11:50
  • $\begingroup$ That is great! All the best! $\endgroup$ – Toros91 Jan 15 at 11:58
  • $\begingroup$ Post training and validation once a model is selected and deployed into practice, does it consider (learn from) further incoming data and keep itself maturing? $\endgroup$ – ranit.b Jan 15 at 16:49
  • $\begingroup$ Generally it depends on the model, for instance if we are talking about random forest then you have to train the mode periodically to improve itself like once in a week or month based on the severity. $\endgroup$ – Toros91 Jan 16 at 0:10

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.