ARMS: an automatic knowledge engineering tool for learning action models for AI planning

2007 ◽  
Vol 22 (2) ◽  
pp. 135-152 ◽  
Author(s):  
KANGHENG WU ◽  
QIANG YANG ◽  
YUNFEI JIANG
Keyword(s):  
Knowledge Engineering ◽  
Learning System ◽  
Propositional Satisfiability ◽  
Ai Planning ◽  
Model Learning ◽  
Frequent Sets ◽  
Action Model ◽  
Action Models ◽  
Definition Of ◽  
Modelling System

AbstractWe present an action model learning system known as ARMS (Action-Relation Modelling System) for automatically discovering action models from a set of successfully observed plans. Current artificial intelligence (AI) planners show impressive performance in many real world and artificial domains, but they all require the definition of an action model. ARMS is aimed at automatically learning action models from observed example plans, where each example plan is a sequence of action traces. These action models can then be used by the human editors to refine. The expectation is that this system will lessen the burden of the human editors in designing action models from scratch. In this paper, we describe the ARMS in detail. To learn action models, ARMS gathers knowledge on the statistical distribution of frequent sets of actions in the example plans. It then builds a weighted propositional satisfiability (weighted SAT) problem and solves it using a weighted MAXSAT solver. Furthermore, we show empirical evidence that ARMS can indeed learn a good approximation of the finally action models effectively.

scholarly journals Learning Partially Observable Deterministic Action Models

2008 ◽  
Vol 33 ◽  
pp. 349-402 ◽  
Author(s):  
E. Amir ◽  
A. Chang
Keyword(s):  
Reinforcement Learning ◽  
Logical Structure ◽  
Exact Algorithms ◽  
Ai Planning ◽  
Partial Observations ◽  
Domain Structures ◽  
Action Model ◽  
Action Models ◽  
Partially Observable ◽  
Over Time

We present exact algorithms for identifying deterministic-actions' effects and preconditions in dynamic partially observable domains. They apply when one does not know the action model(the way actions affect the world) of a domain and must learn it from partial observations over time. Such scenarios are common in real world applications. They are challenging for AI tasks because traditional domain structures that underly tractability (e.g., conditional independence) fail there (e.g., world features become correlated). Our work departs from traditional assumptions about partial observations and action models. In particular, it focuses on problems in which actions are deterministic of simple logical structure and observation models have all features observed with some frequency. We yield tractable algorithms for the modified problem for such domains. Our algorithms take sequences of partial observations over time as input, and output deterministic action models that could have lead to those observations. The algorithms output all or one of those models (depending on our choice), and are exact in that no model is misclassified given the observations. Our algorithms take polynomial time in the number of time steps and state features for some traditional action classes examined in the AI-planning literature, e.g., STRIPS actions. In contrast, traditional approaches for HMMs and Reinforcement Learning are inexact and exponentially intractable for such domains. Our experiments verify the theoretical tractability guarantees, and show that we identify action models exactly. Several applications in planning, autonomous exploration, and adventure-game playing already use these results. They are also promising for probabilistic settings, partially observable reinforcement learning, and diagnosis.

Learning Planning Domain Descriptions in RDDL

2015 ◽  
Vol 24 (03) ◽  
pp. 1550002 ◽  
Author(s):  
Dongning Rao ◽  
Zhihua Jiang
Keyword(s):  
Inductive Learning ◽  
Error Rates ◽  
Description Language ◽  
Model Learning ◽  
Learning Technique ◽  
Finite State ◽  
Action Model ◽  
Action Models ◽  
Domain Description ◽  
Machine Analysis

Recently, there is increasing interest in action model learning. However, most previous studies focused on learning effect-based action models. On the other hand, a rule-based planning domain description language was proposed in the latest planning competition. That is the Relational Dynamic Influence Diagram Language (RDDL). It uses rules to describe transitions instead of action models. In this paper, we build a system to learn planning domain descriptions in the RDDL. There are three major parts of an RDDL domain description: constraints, transitions and rewards. We first take advantage of the finite state machine analysis to identify constraints. Then, we employ the inductive learning technique to learn transitions. At last, we use regression to fix rewards. The evaluation was performed on benchmarks from planning competitions. It showed that our system can learn domain descriptions in the RDDL with low error rates. Moreover, our system is developed based on classical approaches. It implicates that the RDDL roots in previous planning languages. Therefore, more classical approaches could be useful in the RDDL domains.

Cost-Sensitive Action Model Learning

2016 ◽  
Vol 24 (02) ◽  
pp. 167-193 ◽  
Author(s):  
Dongning Rao ◽  
Zhihua Jiang
Keyword(s):  
Online Learning ◽  
Real World ◽  
Learning Algorithm ◽  
High Rate ◽  
Model Learning ◽  
Total Cost ◽  
Cost Constraints ◽  
Action Model ◽  
Definition Of ◽  
The Cost

Action model learning can relieve people from writing planning domain descriptions from scratch. Real-world learners need to be sensitive to all kinds of expenses which it will spend in the learning. However, most of previous studies in this research line only considered the running time as the learning cost. In real-world applications, we will spend extra expense when we carry out actions or get observations, particularly for online learning. The learning algorithm should apply more techniques for saving the total cost when keeping a high rate of accuracy. The cost of carrying out actions and getting observations is the dominated expense in online learning. Therefore, we design a cost-sensitive algorithm to learn action models under partial observability. It combines three techniques to lessen the total cost: constraints, filtering and active learning. These techniques are used in observation reduction in action model learning. First, the algorithm uses constraints to confine the observation space. Second, it removes unnecessary observations by belief state filtering. Third, it actively picks up observations based on the results of the previous two techniques. This paper also designs strategies to reduce the amount of plan steps used in the learning. We performed experiments on some benchmark domains. It shows two results. For one thing, the learning accuracy is high in most cases. For the other, the algorithm dramatically reduces the total cost according to the definition of cost in this paper. Therefore, it is significant for real-world learners, especially, when long plans are unavailable or observations are expensive.

scholarly journals Forensic Engineering And The Scientific Method

2009 ◽  
Vol 26 (1) ◽  
Author(s):  
Laura L. Liptai
Keyword(s):  
Scientific Method ◽  
Knowledge Engineering ◽  
Analytical Framework ◽  
Education Experience ◽  
Forensic Engineering ◽  
Investigative Technique ◽  
Ultimate Objective ◽  
Definition Of ◽  
Prevention Methods ◽  
Engineering Investigation

The Scientific Method Is Utilized In Order To Understand The Relationship Among Observations Of Physical Phenomena, While Minimizing The Influence Of Human Bias And Maximizing Objectivity. Specific Procedures For The Application Of The Scientific Method Vary From One Field Of Science To Another, But The Investigative Technique Universally Provides For An Analytical Framework To Acquire, Collect And/Or Integrate Knowledge. Engineering Forensics Involves The Analysis Of The Parameters Or Cause(S) Of Incidents Or Failures And/Or Hypothetical Prevention Methods. Engineering Analysis Of Forensic Problems Is A Multifaceted, Multidisciplinary Pursuit That Is Often Wide In Scope. Forensic Engineering Generally Applies Existing Science In Conjunction With The Knowledge, Education, Experience, Training And Skill Of The Practitioner To Seek Solution(S). The Scientific Method, Including Definition Of A Null Hypothesis, Is Rarely Utilized In Forensics As New Science Is Rarely Required. A Forensic Engineering Investigation Typically Involves The Application Of Long Established Science (Newtons Laws, For Example). Forensic Engineering Encompasses The Systematic Search For Knowledge Necessitating The Observation And Definition Of A Problem; The Collection Of Data Through Observation, Research, Experimentation And/Or Calculation; The Analysis Of Data; And The Development And Evaluation Of Findings And Opinions. The Ultimate Objective Of A Forensic Engineering Investigation Is Uncompromised Data Collection And Systematically Considered, Iteratively Derived And Objectively Balanced Conclusions.

Knowledge-Intensive Evolutionary Algorithms for Solving a Healthcare Fleet Optimization Problem

2018 ◽  
pp. 192-223
Author(s):  
Carlos Adrian Catania ◽  
Cecilia Zanni-Merk ◽  
François de Bertrand de Beuvron ◽  
Pierre Collet
Keyword(s):  
Evolutionary Algorithms ◽  
Knowledge Engineering ◽  
Domain Ontology ◽  
Real Problem ◽  
Fitness Functions ◽  
Evaluation Functions ◽  
Mutation Operators ◽  
Knowledge Intensive ◽  
Definition Of ◽  
Crossover And Mutation

In this chapter, the authors show how knowledge engineering techniques can be used to guide the definition of evolutionary algorithms (EA) for problems involving a large amount of structured data, through the resolution of a real problem. Various representations of the fitness functions, the genome, and mutation/crossover operators adapted to different types of problems (routing, scheduling, etc.) have been proposed in the literature. However, real problems including specific constraints (legal restrictions, specific usages, etc.) are often overlooked by the proposed generic models. To ensure that these constraints are effectively considered, the authors propose a methodology based on the structuring of the conceptual model underlying the problem, as a labelled domain ontology suitable for optimization by EA. The authors show that a precise definition of the knowledge model with a labelled domain ontology can be used to describe the chromosome, the evaluation functions, and the crossover and mutation operators. The authors show the details for a real implementation and some experimental results.

MedISys

2011 ◽  
pp. 131-142 ◽  
Author(s):  
Jens P. Linge ◽  
Ralf Steinberger ◽  
Flavio Fuart ◽  
Stefano Bucci ◽  
Jenya Belyaeva ◽  
...  
Keyword(s):  
Surveillance System ◽  
Medical Information ◽  
Learning System ◽  
Event Extraction ◽  
Medical Information System ◽  
News Sources ◽  
Fully Automatic ◽  
Definition Of ◽  
Public Health Surveillance System ◽  
The University

The Medical Information System (MedISys) is a fully automatic 24/7 public health surveillance system monitoring human and animal infectious diseases and chemical, biological, radiological and nuclear (CBRN) threats in open-source media. In this article, we explain the technology behind MedISys, describing the processing chain from the definition of news sources, scraping and grabbing articles from the internet, text mining, event extraction with the Pattern-based Understanding and Learning System (PULS, developed by the University of Helsinki), news clustering and alerting, to the display of results. The web interface and service applications are shown from a user’s perspective. Users can display world maps in which event locations are highlighted as well as statistics on the reporting about diseases, countries and combinations thereof and can apply filters for language, disease or location or filters with orthogonal categories, e.g. outbreaks, via their browser. Specific entities such as persons, organizations and locations are identified automatically.

Semantics Enhancing Knowledge Discovery and Ontology Engineering Using Mining Techniques

2011 ◽  
pp. 163-188
Author(s):  
Elena Irina Neaga
Keyword(s):  
Knowledge Discovery ◽  
Knowledge Engineering ◽  
Reference Model ◽  
Distributed Processing ◽  
Ontology Engineering ◽  
Software Architectures ◽  
Model Driven ◽  
Architectural Models ◽  
Definition Of ◽  
Comprehensive Framework

This chapter deals with a roadmap on the bidirectional interaction and support between knowledge discovery (Kd) processes and ontology engineering (Onto) mainly directed to provide refined models using common methodologies. This approach provides a holistic literature review required for the further definition of a comprehensive framework and an associated meta-methodology (Kd4onto4dm) based on the existing theories, paradigms, and practices regarding knowledge discovery and ontology engineering as well as closely related areas such as knowledge engineering, machine/ontology learning, standardization issues and architectural models. The suggested framework may adhere to the Iso-reference model for open distributed processing and Omg-model-driven architecture, and associated dedicated software architectures should be defined.

Methodical system of masters students written academic discourse education with group blog using

2020 ◽  
pp. 35-45
Author(s):  
Ivan S. Dronov
Keyword(s):  
Higher Education ◽  
Social Order ◽  
Cognitive Activity ◽  
Academic Discourse ◽  
Educational Process ◽  
Learning System ◽  
Information And Communications Technologies ◽  
Educational Standard ◽  
Target Definition ◽  
Definition Of

A characteristic feature of the vast majority of activity spheres in the modern education system is their systematization. The latest higher education standards center the teacher’s activity around the system and activity approach. That causes the need for its application in the use of information and communications technologies as a means of teaching a foreign language. We give the author’s definition of the term “methodical learning system”, which is interpreted as a component-related system of learning conditions, realizing its methodological potential through the interaction of participants in the educational process aimed at achieving the goal. This system includes five consecutive blocks, including: 1) prerequisites (modern requirements of the Federal Educational Standard of Higher Education), social order for training in the field of written academic discourse, contradictions between the linguodidactic potential of the study group blog, the lack of practical methods of teaching written academic discourse; 2) block target definition (goals and objectives of training); 3) theoretical justification block (list of principles of teaching written academic discourse, approaches to teaching written academic discourse); 4) functional block (organizational and pedagogical conditions, contents and methods of training, means and organizational forms of training); 5) evaluation block system operation results (components of evaluation of educational and cognitive activity, the learning process result). We note methodical system training integrity on the basis of property of integrativity promoting optimization of all components which general interaction is directed on achievement of the training purpose. The goal, for its part, is a system-forming element.

scholarly journals Preface to special issue on planning and scheduling

2010 ◽  
Vol 25 (3) ◽  
pp. 247-248
Author(s):  
Roman Barták ◽  
Amedeo Cesta ◽  
Lee McCluskey ◽  
Miguel A. Salido
Keyword(s):  
Constraint Satisfaction ◽  
Real World ◽  
Knowledge Engineering ◽  
Real Life ◽  
Ai Planning ◽  
Planning And Scheduling ◽  
Useful Knowledge ◽  
State Dependent ◽  
Life Problems ◽  
Modelling Techniques

AbstractPlanning, scheduling and constraint satisfaction are important areas in artificial intelligence (AI) with broad practical applicability. Many real-world problems can be formulated as AI planning and scheduling (P&S) problems, where resources must be allocated to optimize overall performance objectives. Frequently, solving these problems requires an adequate mixture of planning, scheduling and resource allocation to competing goal activities over time in the presence of complex state-dependent constraints. Constraint satisfaction plays an important role in solving such real-life problems, and integrated techniques that manage P&S with constraint satisfaction are particularly useful. Knowledge engineering supports the solution of such problems by providing adequate modelling techniques and knowledge extraction techniques for improving the performance of planners and schedulers. Briefly speaking, knowledge engineering tools serve as a bridge between the real world and P&S systems.

Sign in / Sign up

Export Citation Format

Share Document