A Programming Language With a POMDP Inside

TL;DR

POAPS is a new programming language and compiler that simplifies the creation of POMDP models for non-experts, enabling adaptive decision-making in complex, partially observable environments like crowdsourcing.

Contribution

It introduces an expressive Lisp-based language with constructs for dynamic optimization, abstracting POMDP complexities for non-expert users.

Findings

Successfully models crowdsourcing tasks with POAPS

Demonstrates ease of use and expressiveness of the language

Shows POAPS can generate effective POMDPs for decision-making

Abstract

We present POAPS, a novel planning system for defining Partially Observable Markov Decision Processes (POMDPs) that abstracts away from POMDP details for the benefit of non-expert practitioners. POAPS includes an expressive adaptive programming language based on Lisp that has constructs for choice points that can be dynamically optimized. Non-experts can use our language to write adaptive programs that have partially observable components without needing to specify belief/hidden states or reason about probabilities. POAPS is also a compiler that defines and performs the transformation of any program written in our language into a POMDP with control knowledge. We demonstrate the generality and power of POAPS in the rapidly growing domain of human computation by describing its expressiveness and simplicity by writing several POAPS programs for common crowdsourcing tasks.