Welcome to the homepage of the HUman-in-the-Loop Computation (HULC) project (pronounced like "HULK")
at the PIKE research group of Penn State, USA.

With the surge of "crowdsourcing" as an alternative computing paradigm to solve non-trivial tasks that are difficult for machines but easy for humans, new opportunities to put "human" in the loop of diverse computational tasks arise.

In particular, in the HULC project, we are interested in solving some computationally-challenging tasks with the use of crowdsourcing arising in Databases, Data Mining, Information Retrieval, Recommender Systems, and WWW fields. Some of the questions that we are addressing in this project include:
  • How to simulate fundamental (e.g., max, sort, join) and advanced (e.g., similar-to, top-k, skyline, order statistic, merge/purge) data processing operations using both humans and machines? What are the theoretical characterizations in optimizing such operations with respect to budget, quality, and latency?
  • When and how to split computationally challenging tasks between humans and machines? When and how to combine tasks done separately by humans and machines?
  • How to design and build a system, say CMS (Crowdsourcing Management System),  to manage the micro-tasks and their complex interactions in crowdsourcing, similar to the way that DBMS manages data?