Searching for surprise

Inspired by the notion of surprise for unconventional discovery in computational creativity, we introduce a general search algorithm we name surprise search. Surprise search is grounded in the divergent search paradigm and is fabricated within the principles of metaheuristic (evolutionary) search. The algorithm mimics the self-surprise cognitive process of creativity and equips computational creators with the ability to search for outcomes that deviate from the algorithm’s expected behavior. The predictive model of expected outcomes is based on historical trails of where the search has been and some local information about the search space. We showcase the basic steps of the algorithm via a problem solving (maze navigation) and a generative art task. What distinguishes surprise search from other forms of divergent search, such as the search for novelty, is its ability to diverge not from earlier and seen outcomes but rather from predicted and unseen points in the creative domain considered.This work has been supported in part by the FP7 Marie Curie CIG project AutoGameDesign (project no: 630665).peer-reviewe