SBIR-STTR Award

This Small Business Innovation Research Phase I project will investigate the potential of using call log data to assist the telecommunications industry in better serving its customers. Call logs (records of who called whom) can be viewed as social networks, with phone numbers representing vertices and phone calls representing edges. Understanding a customer's social network can potentially provide a much better understanding of their behavior and preference. The technical challenge of this project is both statistical and computational. As the telecommunications industry serves a very large number of customers, their data sets are massive. In the single month, for example, a major telecommunications provider logged 12 billion phone calls made between 250 million phone numbers. The technical objectives are: 1) to further develop s nascent computational platform for extremely large-scale network analysis, and 2) to validate algorithms and procedures, which quantify the effectiveness of operator marketing campaigns. The goal is to create s software-and-services product offering designed to leverage telecom companies' own call logs to help them better value, serve and retain their customers. It is believed that the telecommunications industry has overlooked the richest data in their possession: the call logs themselves. A telecommunications operator has information not just on individuals, but on their calling behavior, their communities, and their communities' calling behavior. A single major operator typically serves over 15 million customers; given the sheer number of subscribers, operators must rely on statistical analysis to monitor customer satisfaction and to anticipate customer needs. Enhancing this knowledge will allow them to optimize their product marketing, to improve their customer care strategies, to more efficiently use their advertising budget, and to anticipate "churn," the cessation of service. The mobile phone market currently reaches over 4 billion subscribers globally. While the past decade has seen significant growth, most markets have reached saturation; mobile phone operators now are shifting their focus from growth to efficient customer retention strategies. This shift in strategy presents an opportunity to apply data mining techniques to call logs, a rich resource that to date has remained unused by the industry.This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).

This Small Business Innovation Research (SBIR) Phase II project seeks to create a new, innovative system to manage a highly-scalable, geographically-distributed labor force through wireless technology - what is refered to as " mobile crowdsourcing." The plunging cost of handsets and the introduction of prepaid call plans have allowed individuals throughout the world to have the ability to communicate and transact electronically. This project will create the infrastructure needed to provide wireless subscribers the ability to do work and earn money - leveraging today's mobile phone's ability to send, receive and display images, audio files and text. The system will: deconstruct a client's work into "micro-tasks;" preferentially route micro-tasks to individuals most likely able to complete them; statistically analyze completed work across individual responses to automatically reach a decision on when work is complete, and who has provided the most useful input; compensate workers in proportion to the value they have added; and, finally, reconstruct the completed task for the client, with a statistical assurance the work has been accomplished correctly. The first application of this system will be for the business process outsourcing (BPO) industry. The company will integrate with several mobile carriers in Africa and South America to allow subscribers direct access to transactional BPO tasks including transcription, translation and text categorization. Communicating with workers directly through phones and emphasizing quality control on work, rather than worker will enable users to perform tasks when they want, where they want, and as they want. Automated compensation through existing mobile payment and airtime transfer systems will allow for much lower overhead costs. In addition to cost savings, however, clients who use this system to complete work will also have the benefits of: increased security (no one worker will be able to see an entire document or hear an entire audio recording), access to a scalable workforce (when "spikes" of work come through, labor can be seamlessly scaled up), and potential for very fast turnaround on work (micro-tasks can be done in parallel by many individuals, greatly reducing total time to complete a workload). Additional applications of the mobile crowdsourcing platform include data gathering related to local content and surveys, productivity tools for auditors, and mass reporting abilities following disaster-related events