Browse Prior Art Database

Extracting And Suggesting Business Processes Out Of Organization Communication Interaction Disclosure Number: IPCOM000240365D
Publication Date: 2015-Jan-27
Document File: 5 page(s) / 123K

Publishing Venue

The Prior Art Database


Disclosed is a method and system for extracting and suggesting business processes out of organization communication interaction.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 33% of the total text.

Page 01 of 5

Extracting And Suggesting Business Processes Out Of Organization Communication Interaction

Disclosed is a method for extracting and suggesting business processes received from organization communication. The method processes logs and content from one or more communication sources and detects resource interaction patterns to detect and define processes.

Fig. 1 illustrates the method disclosed herein .

Fig. 2 illustrates details of content parsing 102 .

Referring to Fig. 1, single or consolidated communication databases 101 are illustrated, according to embodiments of the present invention. Mail Database 101 is used as an example. At 102 content available in database 101 is parsed into forms suited for processing within subsequent steps. At 103 patterns are detected from the flow of parsed instances, and resources interaction is represented within the parsed instances. The result, at 104, includes a list of common patterns of interaction, or workflows. These common patterns are put in a standard form at 105 for comparison with existing processes from Processes Database or equivalent storage units 108. The comparison is done at 107, new cases are added at 109 to Processes Database (108, and old cases similar to already known processes are discarded at 110.

Fig. 2 shows more details for content parsing 102 of Fig. 1. Content parsing 102


Page 02 of 5

includes processing individual mail instances as input 201 to extract a sender for each instance at 202. At 203, the disclosed method replaces each sender with a role from Employees Database 210, which stores employer information for all company resources.

Next a list of receivers is extracted at 204, which may include extracting receivers from the "To" field, the "CC" field, or both. This may depend on whether the sender used the "Reply All" button.

For the extracted receivers, the Employees Database is used again at 205 to add roles such as Country CFO or HR Coordinator. A receiver's role may include a relation to the sender, such as the sender's manager or the sender's department director. Also, the method can consider functional accounts that are used to submit or communicate for easily identifying the role.

At 206 communication title and content is parsed to extract important and significant information. Common techniques may be applied based on machine learning or handcrafted rules as part of speech tagging, co-reference and reference extraction, stop words removal, and other techniques to provide information about the purpose of the communication and its attributes. Content is then produced in the form of structured format for further processing such as instances comparison or content normalization, for example. Also, if there are attachments to the communication instance, these are parsed at 208. Parsing may include opening and checking the content of files such as text documents, word documents, presentations, portable document formats, or others. Scanned images maybe candidate...