h-IQ: Human Intelligence for Quality of Service Delivery Data

Service delivery centers are extremely dynamic environments in which large numbers of globally distributed system administrators (SAs) manage a vast number of IT systems on behalf of customers. SAs are under significant time pressure to efficiently resolve incoming customer requests, and may fall far short of accurately capturing the intricacies of technical problems, affecting the quality of ticket data. At the same time, various data stores and warehouses aggregating business insights about operations are only as reliable as their sources. Verifying such large data sets is a laborious and expensive task. In this paper we propose system h-IQ, which embeds a grading schema and an active learning mechanism, to identify most uncertain samples of data, and most suitable human expert(s) to validate them. Expert qualification is established based on server access logs and past tickets completed. We present the system and discuss the results of ticket data assessment process.

[1]  Vijay K. Naik,et al.  Managing Enterprise IT Systems Using Online Communities , 2011, 2011 IEEE International Conference on Services Computing.

[2]  Jim Laredo,et al.  Service for Crowd-Driven Gathering of Non-Discoverable Knowledge , 2011, ICSOC Workshops.

[3]  Abraham Bernstein,et al.  From active towards InterActive learning: using consideration information to improve labeling correctness , 2009, HCOMP '09.

[4]  Jim Laredo,et al.  PeopleCloud Service for Enterprise Crowdsourcing , 2010, 2010 IEEE International Conference on Services Computing.

[5]  G. G. Stokes.  "J." , 1890, The New Yale Book of Quotations.

[6]  Daren C. Brabham.  Crowdsourcing as a Model for Problem Solving , 2008 .

[7]  Jane Yung-jen Hsu,et al.  PhotoSlap: A Multi-player Online Game for Semantic Annotation , 2007, AAAI.

[8]  Luis von Ahn.  Games with a Purpose , 2006, Computer.

[9]  Hans Thies,et al.  Quality Assurance for Human-Based Electronic Services: A Decision Matrix for Choosing the Right Approach , 2010, ICWE Workshops.

[10]  Panagiotis G. Ipeirotis,et al.  Get another label? improving data quality and data mining using multiple, noisy labelers , 2008, KDD.

[11]  Jennifer G. Dy,et al.  Active Learning from Crowds , 2011, ICML.

[12]  Osamuyimen Stewart,et al.  Crowdsourcing participation inequality: a SCOUT model for the enterprise domain , 2010, HCOMP '10.

[13]  Yixin Diao,et al.  Optimizing Change Request Scheduling in IT Service Management , 2008, 2008 IEEE International Conference on Services Computing.

[14]  Osamuyimen Stewart,et al.  Designing crowdsourcing community for the enterprise , 2009, HCOMP '09.

[15]  David A. Forsyth,et al.  Utility data annotation with Amazon Mechanical Turk , 2008, 2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops.

[16]  Larisa Shwartz,et al.  Towards an optimized model of incident ticket correlation , 2009, 2009 IFIP/IEEE International Symposium on Integrated Network Management.