R Code for Incident Ticket Management

text_mining
classification
multi-class

#1

Hi AV Guys

I wan the resources R code, links for R code, ideas for R code for Analytics for incident ticket management in following areas:

I am mainly concern about these R script of following areas on incident ticket management :

i) Comparison questions
a. Are there any significant differences in the
ticket volume across different applications or
groups?
b. How does the actual ticket volume distribution
across priorities or severities conform to the
original assumptions for the portfolio?
c. Are there any significant differences in
resolution times across different ticket
categories or groups?
d. How do ticket resolutions meet the service
level agreement (SLA) requirements?

ii) Trending Questions
a. How does the ticket volume change over time?
Are there any significant patterns or trends?
b. How does the ticket resolution time change
over time?

iii). Forecasting questions
a. What are the anticipated ticket volumes for the
next time period (month, quarter, etc.)?

iv.) Ticket technique structure and resolution strategy
questions
a. What are the typical problem symptoms or
major problematic areas that contribute to
most of the tickets?
b. What are the common solutions to tickets of
different categories?
c. Is it possible to automate ticket resolution?

more reference in paper " Incident Ticket Analytics for IT Application Management Services"

I need the reply ASAP as I am running out of time for project delivery

PS: I have got various papers and blogs but all are theoretical i.e. without any R code

Warm Regards
Manish Sharma


#2

@manishceeri,

These are some articles in R that can help you with your project -

  1. For text mining in R refer,
    a. Building word cloud with R https://www.analyticsvidhya.com/blog/2014/05/build-word-cloud-text-mining-tools/
    b. Sentiment analysis of Twitter using R https://www.analyticsvidhya.com/blog/2017/03/measuring-audience-sentiments-about-movies-using-twitter-and-text-analytics/

  2. For classification refer,
    a. Text Classification using FastText https://www.analyticsvidhya.com/blog/2017/07/word-representations-text-classification-using-fasttext-nlp-facebook/

Regards,
Sanad :slight_smile:


#3

Hi

Thanks for prompt reply.

I have already gone through these links and others. Now i want to know the R script flow pertaining to " Incident Management ticket analysis".

I am mainly concern about these R script of following areas on ITSM:

i) Comparison questions
a. Are there any significant differences in the
ticket volume across different applications or
groups?
b. How does the actual ticket volume distribution
across priorities or severities conform to the
original assumptions for the portfolio?
c. Are there any significant differences in
resolution times across different ticket
categories or groups?
d. How do ticket resolutions meet the service
level agreement (SLA) requirements?

ii) Trending Questions
a. How does the ticket volume change over time?
Are there any significant patterns or trends?
b. How does the ticket resolution time change
over time?

iii). Forecasting questions
a. What are the anticipated ticket volumes for the
next time period (month, quarter, etc.)?

iv.) Ticket technique structure and resolution strategy
questions
a. What are the typical problem symptoms or
major problematic areas that contribute to
most of the tickets?
b. What are the common solutions to tickets of
different categories?
c. Is it possible to automate ticket resolution?

more reference in paper " Incident Ticket Analytics for IT Application Management Services"

I am also editing the question.

Warm Regards
Manish Sharma


#4

hi manish
can you just share the script used for above …Incident Management ticket analysis .if you were able create one


#5

@vinosamvarghese

I am working to figure out about how to implement R code, that is why i have posted the question here, hope guys like @vopani could take a bit time from their busy schedule and throw some light into this.


#6

Same …i am too working on it too…keep updated


#7

Hi

I have the data set in .csv format and in this data set I have the following columns:

Created
1/1/2017 1:03
1/1/2017 3:42
1/2/2017 10:36
1/2/2017 11:09

and

Close Time
1/1/2017 2:54
1/1/2017 4:14
1/9/2017 15:19
1/2/2017 14:48
1/2/2017 12:35
1/2/2017 11:58
1/2/2017 13:07
1/3/2017 17:30
1/2/2017 13:48

Now as visible these data are the daily recordings i.e of everyday per month for the year 2017.

I now want to know

i) the ideas that I can formulate using these columns ( like how can i use time series)

ii) how to combine / merge these daily reading month wise i.e how to make a month columns having the readings of that month only so that I can do visualization month wise.

iii) one column is ticket description, how to extract this column only for topic modelling.

I want to implement these in R only.

Regards


#8

HI Guys

I have performed LDA; topic modelling on ticket description using harmonic mean for best value of K, i have visualize the topics thus formed, but i find that I have many words repeating in more than one topic.

how to visualize these correlated topics and their words?
how to delete these correlated words or how to deal with them?


#9

Not sure if I am too late here, but i have a script that walks though an Incident management dataset from a tool called ServicNow, it creates approx. 17 visuals (bar charts, line charts, word clouds etc) it does text mining shows you the most common X issues, categories, configuration items, assignment groups, top phrases, etc. What would be interested in looking into is the changes of things across time, I personally feel that is the golden “goose egg”. let me know if this would be of interest.