Browse Prior Art Database

System and Method to discover virtual teams by analyzing call log of instant messaging systems Disclosure Number: IPCOM000109154D
Original Publication Date: 2005-Mar-23
Included in the Prior Art Database: 2005-Mar-23
Document File: 1 page(s) / 26K

Publishing Venue



Customers are increasingly investing in knowledge management solutions. However, there is a performance gap because these solutions focus on explicitly assimilated digital information and miss ~50% of the information in the general office which transpires using other modalities such as telephone, meetings and paper. Until a few years ago, instant messaging was a multitasking activity reserved for agile adolescents eager to chat with as many pals as possible at once -- often with one hand on the keyboard, and the other on the telephone. In recent times, corporate users have discovered that instant messaging helps exchange small but often critical details back and forth in a rocket-like speed, delivering yet another shortcut to the already fast-track world of electronic messaging. A typical e-mail thread may cycle three to five days towards completion. However, the same participants in that e-mail discussion when empowered with instant messaging, can make decisions in real time, and perhaps shorten that thread from five days to five minutes. This has resulted in instant messaging becoming a critical business productivity tool in many enterprises. Therefore, corporate instant message logs are an example of a naturally occuring modality in the general office environment from which organizational, collaborative and knowledge assets can be extracted. In this context, the goal is to transform a corporate instant message corpus into a useful knowledge asset that can be used for several purposes such as organizational change management, to improve project management etc.In today's enterprises people frequently collaborate with others that are not part of their "official" team, but organize spontaneously in "virtual-teams". Keeping track of such teams is made difficult by their very flexibility and volatility. One way is to ask people to list the members of their virtual teams, but this is slow, costly and difficult to keep up-to-date. Another way is to reorganize the company around the virtual teams, but this is practical only when the virtual teams are stable enough and long lasting, which is rare in modern organizations. This disclosure proposes a set of algorithms to automatically derive such knowledge assets from an Instant Message corpus. Current solutions related to instant message management have more to do with capture, storage and retrieval of the messages. The analysis of the data to organize into useful categories is either manually done (such as Lotus BusinessChat product offering) or is typically not done. IBM is in the enviable position of consolidating innovative analytical approaches on an integrated enterprise IT platform to deliver compelling solutions.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 63% of the total text.

Page 1 of 1

System and Method to discover virtual teams by analyzing call log of instant messaging systems

The idea of this invention is to take advantage of the communication already taking place among the members of the virtual teams. Computer-based communication media, email and instant messaging, can log the names of the people involved in a conversation. It is possible to analyze such logs to automatically discover virtual teams. The advantage provided by our system is that the detection of virtual teams is automated and can be easily repeated as frequently as necessary, without need for employees to spend time updating their records.

The idea stems from the observation that members of virtual teams tend to communicate with all the other members on their team. If we represent the communication log as a graph where the nodes are the people and the arcs the conversation that have taken place, the virtual team will look like highly interconnected sets of nodes with few connections to nodes outside the virtual team. The algorithms we propose aims to find such highly interconnected subsets of people.

The preferred implementation of the invention is as a search service accessed via a web-based interface. The search service will get the log directly from the sametime server and produce a list of possible virtual teams that can be searched by the user of the system using the web interface. In order to track the virtual teams in their evolution the system will analyze the sameti...