Reduce Artificial Intelligence Planning Effort by using Map-Reduce Paradigm

While several approaches have been developed to enhance the efficiency of hierarchical Artificial Intelligence planning (AI-planning), complex problems in AI-planning are challenging to overcome. To find a solution plan, the hierarchical planner produces a huge search space that may be infinite. A p...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:International journal of innovative technology and exploring engineering 2021-05, Vol.10 (7), p.24-32
Hauptverfasser: Elkawkagy, Mohamed, Elbeh, Heba
Format: Artikel
Sprache:eng
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:While several approaches have been developed to enhance the efficiency of hierarchical Artificial Intelligence planning (AI-planning), complex problems in AI-planning are challenging to overcome. To find a solution plan, the hierarchical planner produces a huge search space that may be infinite. A planner whose small search space is likely to be more efficient than a planner produces a large search space. In this paper, we will present a new approach to integrating hierarchical AI-planning with the map-reduce paradigm. In the mapping part, we will apply the proposed clustering technique to divide the hierarchical planning problem into smaller problems, so-called sub-problems. A pre-processing technique is conducted for each sub-problem to reduce a declarative hierarchical planning domain model and then find an individual solution for each so-called sub-problem sub-plan. In the reduction part, the conflict between sub-plans is resolved to provide a general solution plan to the given hierarchical AI-planning problem. Pre-processing phase helps the planner cut off the hierarchical planning search space for each sub-problem by removing the compulsory literal elements that help the hierarchical planner seek a solution. The proposed approach has been fully implemented successfully, and some experimental results findings will be provided as proof of our approach’s substantial improvement inefficiency.
ISSN:2278-3075
2278-3075
DOI:10.35940/ijitee.G8902.0510721