Overview

Skill Level: Beginner

New to IBM WEX

This recipe gives a simple and multipurpose basis to create a cognitive tool to analyze a ticket list taken from an issue tracking system.

Ingredients

IBM Watson Explorer Content Analytics - Administration Console

IBM Watson Explorer Content Analytics - Content Analytics Miner

Step-by-step

  1. Export data from the issue tracking system

    1.1 Export the data from the issue tracking system.

    1.2 Save the data into csv format.

    1.3 Choose UTF-8 encoding.

  2. Create a new collection

    2.1 Log in into IBM WEX administration console.

    2.2 Create a new collection by pushing the button “Create collection”.

    1-3

    2.3 Choose the name of the collection (i.e. “TicketAnalysis”). The collection type is “Content Analytics Collection“. Then press “OK”.

    2-3

    2.4 You are redirected to the Home Page and see the new collection.

    3-3

  3. Import csv data

    3.1 Within the new collection in the Home Page press the + on the Import bar and choose “CSV import“.

    3.2 Choose the local path of your csv file within the section “CSV file path” and press “Next”.

    3.2

    3.3¬†Encoding character set is “UTF-8”. Check “Read the starting line as a header“. Press “Next”.

    4-3

  4. Create facets

    4.1 In the first column check only the fields you want to analyse. 

    4.2 On the column “Index field name” select New and choose the facet name for the fields you want to analyse (avoid special characters).

    4.3 The “Parametric search” column is “Decimal” for numeric fields and “Date” for dates.

    4.4 “Text sortable” is checked for every field not being a number or a date.

    4.5 Check all the other fields.

    4.6 In “Advanced options” you can decide whether to choose the¬†unique¬†identifier from the existing columns (4.6.1) or to create a new identifier (4.6.2):

    ¬† ¬† ¬† ¬† ¬†4.6.1 In this case check “Generate unique identifier automatically”.

    ¬† ¬† ¬† ¬† ¬†4.6.2 In this case decheck “Generate unique identifier automatically” and select¬†as “Column to use as the document date” the field that will be the key of each row. In the table check the unique field as “Unique identifier”.

    4.7 The “Date format” helps decode the dates in the date fields. Insert the various date formats separated by “;”.

    In this case the formats are “mm/dd/yyyy;m/d/yyyy;mm/d/yyyy;m/dd/yyyy”.

    4.8 Choose the language of the dataset.

    4.9 Press “Next” and then “Finish”.

    5-3

    4.10 Save the current settings by choosing a proper name and pressing “Finish”. This will be helpful in case you’ll want to¬†add¬†a new dataset and upload it reusing the same settings.¬†

    5.2

    4.11 You’ll be redirected to the Home Page. Take a break, wait a moment, refresh the page and you’ll see that the number of documents in the “Parse and Index” section is the same of the number of rows of the csv file.¬†

    6-4

  5. Configure the date facet

    5.1 From your collection press “Action” -> “Settings” -> “Edit collection settings”.

    5.2 Check your date facets under “Index fields used as a date facet in the content analytics miner”.

    7-4

  6. Make the changes effective

    5.1 From the Home Page in your collection press “Deploy analytic resources”.

    5.2 From the Home Page in your collection press “Rebuild index”.

    9

  7. Content Analytics Miner: Facet view

    6.1 Log-in into the Content Analytics Miner.

    6.2 Select your collection from the upper right men√Ļ.

    6.3 Select the “Facet” tab.

    6.4 Here you see:

    ¬† ¬† ¬† ¬† ¬† 6.4.1 The frequency of each state by pressing the facet “State”.

    10-1

    ¬† ¬† ¬† ¬† ¬† 6.4.2 The frequency of each severity by pressing the facet “Severity”.

     11

    6.5¬†If you select¬†a value of your interest for a certain facet you¬†can filter by this value by pressing the lens “Add to query with boolean END”.

    6.6 Press the tab “Document” to see all the issues filtered through the facet value you selected.

     

  8. Content Analytics Miner: Time series view

    7.1 Select the tab “Time series”.

    7.2 The “date facet” is your data field.

    7.3 You see the number of tickets for each year.

    12

    7.4 Select “Time scale” “Month”.

    7.5 You see the number of tickets for each month.

    13

  9. What's next

    8.1 Create an annotator to detect common issues and solutions through the Content Analytics Studio.

    8.2 Create a dashboard.

    14

    8.3 Create a Case Base Reasoning system, in which you can solve a new issue by finding similar issues that occurred in the past and then adapt the old solution to the new problem to solve it.

Join The Discussion