Browse Prior Art Database

Applying map-reduce techniques of unstructured data to enrich workflows for improved tracking

IP.com Disclosure Number: IPCOM000246566D
Publication Date: 2016-Jun-17
Document File: 3 page(s) / 93K

Publishing Venue

The IP.com Prior Art Database

Abstract

Applying map-reduce techniques of unstructured data to enrich workflows so that tracking can be improved.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 53% of the total text.

Page 01 of 3

Applying map-reduce techniques of unstructured data to enrich workflows for improved tracking

Disclosed is a method which will enable map-reduce techniques to be applied to unstructured data to enrich workflows so that tracking can be improved.

Business workflows are predefined by a set of tasks and decision points. The actual execution of a workflow is influenced by the decisions. Companies have defined workflows for their various processes. The workflows typically define a state machine for an item and a state machine is defined by different states. For example, a claim can have these states defined: received, investigated, waiting for confirmation, approved and denied.

    The structure of the workflow can quickly get out of date and new states can be added after the workflow system has been deployed. In addition, some of the states could be optional. An insurance company might decide that every suspicious claim might need to be investigated by an external party. In this case, some new states for the claim need to be added such as: waiting to be sent to the external party, report chased, report received etc. It can often be difficult to update the workflow to reflect the new states.

    In addition to the workflow state, there may be additional information about a specific claim, for example, that is contained in emails, instant messaging conversations, forums, application exchanges, VOIP conversations, etc. For example, an email from someone might say "I have discussed with the customers about X and they said that they will add some comments". In the workflow, the state is "investigated" but its actual state is "waiting for comments from the customers". We'll call the first state the "system state" and the latter the "real state".

    The problem is that the workflows do not reflect the real time states that can be caused by a loosely defined or an out-of-date workflow or a long time transition between two states.

    The idea is to start from the system states coupled with the timestamp associated with each transition from state to state. The system would then find any associated data relating to the particular workflow instance and map them to each particular transition e.g. these emails were exchanged relating to this claim on the transition from "received" to "assigned an evaluator" state.

    The system would attempt to reduce the state of the last transition between two system states, say from "being investigated" to "investigation completed", and use the related information (emails, messaging conversations) associated with this transition to infer the real state.

    The idea here will be described as an embodiment of email software but the same idea can be applied to any workflow system.

    Here are the steps:
1. Identify the defining characteristic of the wor...