Browse Prior Art Database

System for detection of overlapping projects within an organization Disclosure Number: IPCOM000246411D
Publication Date: 2016-Jun-06
Document File: 3 page(s) / 146K

Publishing Venue

The Prior Art Database


In big organizations it happens very often that several teams are working on the same task/project (unaware of that fact). It also happens that there are two almost identical offerings available for customers. Such situation leads to waste of money and confusion across sales teams/customers. It also happens that currently different products/offering have a lot of common items/features in backlog to be implemented in the future (in close future the products will have overlapping functionality). In such case it would desired to merge implementation effort.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 3

System for detection of overlapping projects within an organization
The idea of this article is a system that allows to gather data from different projects and find projects that covers the same functionality or will cover the same functionality in the future . A project may contain different types of data: source code, presentations, schedules, design documents, API specifications, data from a tracking system and version control system, email and chat conversations, forums, audio and video team sessions (statuses, demos), etc. The system gets data from a set of projects and then, based on semantic analysis of their content and analysis of the source code, finds projects with the same domains and overlapping functionality. By analysis of natural language documents and schedules, the system can predict plans for the future and content of future deliverables. The proposed solution has several advantages over current systems:

- Allows to build a semantic model based on artefacts specific for source code: software libraries with their description and API documentation, usage of external software libraries from source code, names of methods, fields and annotations, comments included in the source code, comments included in a tracking system and a version control system.

- Allows to predict a future scope of a project based both on current plans and history of changes of project artefacts. The analysis of historical changes of project assumptions, goals, and plans is also used to compare future scope of projects.

- The system connects and analyses different kinds of project data: documents, schedules, diagrams, source code, audio and video material.

The idea of this disclosure is described in the flowing diagram:

Figure 1. Flowchart of the system

1) First, a user defines input data: data sources and definition of projects.

2) Next, during the Correlation and Dependence Analysis Phase, the data is shaped, analyzed, and a semantic graph is created for each project.

3) The next step is finding correlations between all defined projects and creation of a list


Page 02 of 3

of similar products.

4) SCORING: providing likelihood of correlations and prioritizing the list based on it

The second flow is used to predict content of projects in future and find out which projects derived to similar results.

1) The same as previously, a user defines input data: data sources and definition of projects.

2) The next phase is model training based on provided data related to already finished projects. In order to train the model, a training algorithm like SPSS Modeler, Spark MLLib, or Watson Services can be used.

3) The new model is used to predict future silmilarity of projects.

4) A list of correlated project is...