InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Defect backlog trend analysis using time-series data

IP.com Disclosure Number: IPCOM000018996D
Original Publication Date: 2003-Aug-25
Included in the Prior Art Database: 2003-Aug-25
Document File: 5 page(s) / 128K

Publishing Venue



An algorithm that allows defect backlog trend analysis on any given time period to be conducted on the current defect repository.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 48% of the total text.

Page 1 of 5

Defect backlog trend analysis using time-series data

Disclosed is an algorithm that eliminates the need to maintain snapshot copies of the defect repository. It can be applied to any time interval and allows for the automation of the defect analysis without any manual intervention.

Development teams that want to understand the state of their defect backlog typically maintain snapshots of their defect repositories to allow trend analysis to be conducted over time. The granularity of the analysis is determined by the frequency of the snapshots which is often limited by the size of the repository that must be archived. Also, once the snapshot data is collected the time interval of the displayed data is dictated by the frequency of the snapshots and it can not be changed. Analysis of this data is then conducted using a combination of simple queries, tools and manual steps. The analysis is therefore time-consuming and error prone.

This innovation proposes a new, automated methodology of performing backlog analysis without the need of the multiple snapshots currently used. It can be applied to any time interval, and can be combined with other business specific parameters, such as component, severity, etc. for more detailed analysis.

The algorithm works by calculating the number of defects opened and closed during the software development cycle (in process defects). For the time period of interest, it aggregates the data for specified time intervals.(yearly, monthly, weekly, daily). For each specified time period the number of opened, closed and active defects is computed using the Opendate and Closedate time stamps. The algorithm assumes that the opendate and closedate of a defect are unique and do not change and that the complete defect repository for the project is available (that is, the assumption is there is no "opening" backlog). Queries can be run against any subset period of time.

Note: The algorithm does not address defect state changes (that is, changes in the status of a defect, as recorded through a parameter such as State or Status), or other changes that a defect record may incur while opened.

The algorithm has several steps:

The time interval for the analysis is defined (for example: yearly, monthly, weekly, daily)

Given the time interval, the opendate for each defect is changed to the beginning of the

interval. For instance, if the time interval is "monthly", and the opendate for a particular defect is July 23, the opendate is changed to July 01. The opendate and the number of defects opened for each interval is computed and stored in

temporary Table A. Using the same approach, the closedate and the number of defects closed for each interval is

computed and stored in temporary Table B. A backlog table is created through a "join" of Table A and Table B, based on interval.

If an interval record is missing for Table A after the join, the number of open defects is set to zero for the missing interval







Page 2...