Browse Prior Art Database

A Scientific Approach for Risk Evaluation for Defect Fixing

IP.com Disclosure Number: IPCOM000249747D
Publication Date: 2017-Mar-30
Document File: 4 page(s) / 444K

Publishing Venue

The IP.com Prior Art Database

Related People

Anand MR: AUTHOR [+3]

Abstract

Bringing new change to any module or a feature, in any software product, raise the need for Risk Mitigation, Monitoring and Management (RMMM) as any change can lead to negative Net Promoter Score (NPS). This paper talks about a scientific approach, which can aid in identifying the level of risks involved. Defect Risk Analyzer (DRA) defines heuristics, while applying Data Sciences and machine learning techniques, to reach to a scientific formula to evaluate the same.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 44% of the total text.

A scientific approach for risk evaluation for defect fixing Authored By:

Anand MR Sundip Sharma

Satarupa Pal

ABSTRACT Bringing new change to any module or a feature, in any software product, raise the need for Risk Mitigation, Monitoring and Management (RMMM) as any change can lead to negative Net Promoter Score (NPS). This paper talks about a scientific approach, which can aid in identifying the level of risks involved. Defect Risk Analyzer (DRA) defines heuristics, while applying Data Sciences and machine learning techniques, to reach to a scientific formula to evaluate the same. 1. INTRODUCTION In this paper, we will talk about an algorithm to do risk analysis related to any technical change, proposed in a software. It takes into consideration various parameters like feature usage, impact of changes on product, existing customer-faced issues in that features, lines of code changes, severity of errors, cyclomatic complexity etc.

First phase of DRA involves Data Collection. It collects historical data-sets from various sources like:

1. Defect Tracking System (DTS) 2. Source Control Versioning System (SCV) 3. Live Error Tracking System(LET) 4. Automation Database (ADB) 5. Impact Analyzer (IA) 6. Feature Tracking (FT)

Second phase of DRA involves Data Analysis for Risk Measurement.

Throughout the paper, we will use an example of QuickBooks Desktop (QBDT) which is a product by Intuit. QuickBooks is a financial software which provides business solutions to Small Business Owners.

2. PHASE-1 :: DATA COLLECTION Below are the details on various fields which can be consumed from data collection sources, which will be the contributors for heuristic.

2.1 Data from Defect Tracking System (DTS):

Listed below is the information that can be retrieved

from DTS:

History of the customer faced defects raised in this module/area

Internal defects raised during regression testing

Severity of the defects

Source code change number or change-list Id

Re-open, duplicate and linked defects count etc.

Example : QBDT uses JIRA(a tool for project management, while doing bug, issue and feature tracking) for product management. JIRA contains information of nature –

- Priority (P0, P1, P2) - Severity (Critical, Major, Medium and

Minor) - Issue re-open frequency - Call volume - Frequency (Always, Intermittent etc.) - Impact (High, Medium, Low) etc.

2.2 Data from Source Control Versioning System

(SCV):

Listed below is the information that can be retrieved

from SCV:

 For a particular period, number of file changed for given module/feature

 Lists of changed the files and the line of code changes

 DTS Ids related to the changes done

Example: Perforce is a source control versioning system used in QBDT. From the Perforce we can get following information:

- History of the file modified/added.

- Name of users who modified the file.

2.3 Data from Live Error Tracking System (LET): LET collects data from live error tracking system, which monitors customer-faced defects.

Exa...