Medium 9781601322470

Artificial Intelligence: The 2013 WorldComp International Conference Proceedings

Views: 6828
Ratings: (0)

New research by international contributors on Artificial Intelligence

List price: $99.95

Your Price: $79.96

You Save: 20%

Remix
Remove
 

158 Chapters

Format Buy Remix

Session Agent Technologies and Applications

PDF

 

An Autonomous Agent with an Evolved, Affect-based Decision System

PDF

Int'l Conf. Artificial Intelligence | ICAI'13 |

3

An Autonomous Agent with an Evolved, Affect-based Decision System

Paul G. Joseph1,Haim Levkowitz1

Department of Computer Science, University of Massachusetts, Lowell, MA, USA

1

Abstract - We investigate the modeling of affect with the goal to create an agent that behaves "emotionally" closer to humans.

Our previous work showed evidence that modeling the fundamental emotions of the mammalian brain offered the basis of a flexible decision system for an autonomous agent. Earlier work hard-coded the parameters used to model affect; now, we let the parameters evolve instead. Agents learn dynamically from interactions with their environment and use this learning in a speedy, general-purpose decision system. The agent's interactions with its environment and consequent impact on its wellbeing promise an affect-based approach to semantics. We subject an agent to a “Turing-like test for affect” in which it has to pass a hostile predator to reach its goal. We compare our approach with others; the strict basis of our approach is mainstream neuroscience. Agents independently assign affect valuations that provide a basis for semantic understandings.

 

A Meta-Heuristic Approach for Energy-Efficient Multi-Agent Path Planning

PDF

10

Int'l Conf. Artificial Intelligence | ICAI'13 |

A Meta-Heuristic Approach For Energy-Efficient Multi-Agent Path

Planning

Wenjie Wang, and Wooi-Boon Goh

School of Computer Engineering, Nanyang Technological University, Singapore

Abstract— We address the problem of multi-agent path planning (MAPP) by taking into account the total energy consumption of all agents, as measured by their cumulative distance travelled. A deterministic meta-heuristic technique called guided iterative prioritized planning (GIPP) is proposed to find good solutions with respect to this cumulative cost measure in a crowded environment. GIPP progressively explores a larger search space while incorporating a local optimum detection and resolution scheme to prevent the search process from being trapped in an unfeasible locally optimal solution. Comparative evaluations show that GIPP strikes a good tradeoff between the solution quality and computational time.

Keywords: Guided local search, path planning, multiple agents, meta-heuristic technique

 

An Approach for Norm-Based Behavior Modification in Model-Based Reflex Agents

PDF

Int'l Conf. Artificial Intelligence | ICAI'13 |

17

An Approach for Norm-Based Behavior Modification in Model-Based Reflex Agents

Gustavo A. L. Campos1, Emmanuel S. S. Freire1, Mariela I. Cortés1 and Wamberto W. Vasconcelos2

1

Science Computer Department, State University of Ceará, Fortaleza, Ceará, Brazil

2

Department of Computing Science, University of Aberdeen, Aberdeen, UK

Abstract—Norms in multi-agent systems are used to regulate the behavior of agents, organizations and sub-organizations in their environments during a period of time. Much of the work on norms focuses on the specification and maintenance of normative system. Little work concentrate on the level of individual agents as the impact of norms on the different types of intelligent agent programs, as the internal modifications in the agents´ decision processes and the background information necessary for rationality in an environment governed by rules. This paper is a contribution in this direction to the case of modelbased reflex agent. An approach to extend this type of agent was formalized and validated it in an implementation of a simple world in the

 

Coalition-based Approach for Reach Stackers Routing Problem in Container Terminals

PDF

Int'l Conf. Artificial Intelligence | ICAI'13 |

23

Coalition-based Approach for Reach Stackers Routing Problem in

Container Terminals

H. Bouzidi, M. Kefi and K. Ghédira

SOIE Laboratory, University of Tunis, Tunisia

Abstract— The emergence of containerization had revolutionized the freight maritime transport. The number of the daily handled containers is continuously growing, and port authorities are facing new challenges.

Depending on the organization and the level of automation of the container terminal (CT), different vehicles are used to transport the containers, such as Automated Guided

Vehicles , Reach Stackers, Straddle Carriers. In our case, we focus on using Reach Stackers.

Throughout this paper, we present, formulate and propose a model to solve the Reach Stackers Routing Problem(RSRP) as a Vehicle Routing Problem with Time Windows and Pickup and Delivery (VRPTWPD). First, we draw an analogy between these two problems. Second, we present a state of the art of methods dealing with the vehicle routing problems in CT. Finally, we propose an approximate

 

Adjustable Group Behavior of Agents in Action-based Games

PDF

30

Int'l Conf. Artificial Intelligence | ICAI'13 |

Adjustable Group Behavior of Agents in Action-based Games

Westphal, Keith and Mclaughlan, Brian

Kwestp02@uafortsmith.edu, brian.mclaughlan@uafs.edu

Department of Computer and Information

Sciences

University of Arkansas - Fort Smith

Abstract – Non-player characters (NPCs) within a single-player action-based game have historically been easy to kill and often do not work as a team to overwhelm the players. In addition, they generally do not adjust strategies based on the player’s behavior. This research proposes that these agents should have adjustable behavior parameters that allow them to adapt to the behaviors of their opponents. In particular, this research proposes these NPC’s have an adjustable aggressiveness towards the player giving them an option on whether to hide or to attack. Also, it examines the utility of agents that attempt to attack together and work in a group using simple decision making rather than being individuals. We developed a simulation that gathers information regarding the win percentages of teams of agents that have various behavior combinations versus different types of opponents. Behavior parameters examined include individuality versus teamwork and aggressive posture versus defensive posture. Results showed that certain behavior combinations have significant utility when the behavior of the opposing team is identified.

 

An Agent-Based Intentional Multi-Robot Task Allocation Framework

PDF

,QWURGXFWLRQ

057$ SUREOHP KDV EHFRPH D NH\ UHVHDUFK WRSLF LQ WKH ILHOG

RI GLVWULEXWHG PXOWL�URERW FRRUGLQDWLRQ LQ UHFHQW \HDUV� 7KH

JRDO RI 057$ LV WR FKRRVH RQ WLPH D SURSHU DFWLRQ IRU HYHU\

URERW WR HQVXUH WKDW WKH\ FDQ HIIHFWLYHO\ FRPSOHWH DOO WKH WDVNV�

6XSSRVLQJ WKDW DQ\ LQGLYLGXDO RI WKH V\VWHP FDQQRW FRPSOHWH D

FRPSOLFDWHG SUREOHP� WKLV SUREOHP FDQ EH GHFRPSRVHG WR WKH

VHULHV RI VXE�SUREOHPV RQ VRPH SULQFLSOHV DQG WKHVH VXE�

SUREOHPV FDQ GLVWULEXWHG WR FRUUHVSRQGLQJ LQGLYLGXDO WR VROYH

WKH SUREOHP� 3UHVHQWO\� WKHUH DUH PDQ\ PHWKRGV IRU WDVN

DOORFDWLRQ JURXSHG LQ WKUHH PDLQ DSSURDFKHV� IXOO\ FHQWUDOL]HG�

IXOO\ GLVWULEXWHG DQG K\EULG DSSURDFKHV�

$ FHQWUDOL]HG WDVN DOORFDWLRQ DSSURDFK XVHG IRU PXOWL�URERW

V\VWHP VXIIHUV IURP OLPLWDWLRQV LQ ERWK V\VWHP VFDODELOLW\ DQG

UREXVWQHVV� ,I WKH FHQWUDO URERW IDLOV� WKH ZKROH V\VWHP IDLOV�

2Q WKH RWKHU KDQG� D IXOO\ GLVWULEXWHG V\VWHP SURGXFHV

VXERSWLPDO VROXWLRQ� +\EULG DSSURDFKHV FRPELQH ERWK

FHQWUDOL]HG DQG GLVWULEXWHG DSSURDFKHV FRPELQLQJ WKHLU

DGYDQWDJHV� (DFK LQGLYLGXDO LQ WKH V\VWHP ZRUNV IRU LWV RZQ

 

Potential Functions for Voting Dynamics

PDF

42

Int'l Conf. Artificial Intelligence | ICAI'13 |

Potential Functions for Voting Dynamics

Neelam Gohar1 and Paul W. Goldberg1

1 Department of Computer Science, University of Liverpool, Liverpool, United Kingdom

Abstract— Recent work in the AI literature has studied the properties of voting schemes for performing preference aggregation. An important technical issue that arises is manipulation of voting schemes: a voter may be able to improve the outcome (with respect to his own preferences) by reporting his preferences incorrectly. Meir et al. [11] have studied this phenomenon as a dynamic process in which voters may repeatedly alter their reported preferences until either no further manipulations are available, or else the system goes into a cycle. Here we show how potential functions are useful for showing convergence in variants of this setting, and for obtaining upper bounds on the convergence rate. We focus on plurality voting with weighted voters, and obtain bounds on the lengths of sequences of manipulations, that depend on which types of manipulation are allowed.

 

Applying Knowledge Model to Agent Based Systems

PDF

50

Int'l Conf. Artificial Intelligence | ICAI'13 |

Applying Knowledge Model To Agent Based

Systems

Khaoula ADDAKIRI

Mohamed BAHAJ

Hassan 1 University, FSTS

Department of Mathematics and Computer Science

Khaoula.addakiri@gmail.com

Hassan 1 University, FSTS

Department of Mathematics and Computer Science mohamedbahaj@gmail.com

Abstract— Nowadays, multi-agent system is become promising means for the development of distributed systems, however its disadvantage is that it lacks the interconnection with semantic web such as Ontology Web Language (OWL). In this article, we aim to present a semantic knowledge model of an agent suitable for discrete environments as well as implementation and a use of such model using different softwares

(JENA, JADE, JESS and Protégé) in order to allows interconnection of Agent and Semantic Web technologies which can be used in an agent based application where such interconnection is needed.

Keywords— Multi Agent System ; Web Ontology Language

(OWL); Java Expert System Shell (JESS); SPARQL1.1.

 

Intelligent Agent-based Framework for Project Integration Management

PDF

56

Int'l Conf. Artificial Intelligence | ICAI'13 |

Intelligent Agent-based Framework for Project

Integration Management

Ogechikanma Linda Ihekweaba1, Chukwugoziem Ihekweaba1, Prof. H.C Inyiama2

1

Computer Engineering Dept. Michael Okpara University of Agriculture, Umudike, Abia State Nigeria.

2

Department of Computer Electronics, Nnamadi Azikiwe University, Awka, Nigeria.

Abstract - The dynamism required in the continuous aggregation of the results of all subsidiary planning activities in a project into a consistent document(the project plan), the effective and progressive elaboration of such a document with updates ,the continuous communication requirement during project execution activities coupled with the computational effort required for various ongoing performance evaluation analysis and the constant coordination of changes across the entire elements of a project is definitely overwhelming to the project manager using conventional Project Management

Information System. This paper proposes the exploitation of the dynamism of intelligent agents to support project integration management, a core knowledge area of project management. Using the Belief-Desire-Intention

 

Session - International Workshop on Intelligent Linguistic Technologies - Ilintec'13

PDF

 

Method for Extracting Translation Correspondences from a Parallel Corpus

PDF

Int'l Conf. Artificial Intelligence | ICAI'13 |

65

Method for Extracting Translation Correspondences from a Parallel Corpusi

Yu. Morozova

Institute for Informatics Problems of the Russian Academy of Sciences, Moscow, Russia

Abstract – The research paper deals with actual problems of semantic studies using methods of corpus linguistics. It overviews a new research area - distributional semantics.

The method for extracting translation correspondences from a parallel corpus using context vector spaces is described. The model was tested on a parallel corpus of patent texts in Russian and French. An approximate evaluation of precision is 78%. False positive results can be explained by two main reasons. In the first place, productive syntactic transformations result in translation correspondences containing two words which are semantically equivalent but belong to different parts of speech. In the second place, words are often part of multiword expressions and in such cases a correct single-word translation correspondence cannot be found. We propose to enhance existing models by moving from single lexemes to multi-word expressions.

 

Knowledge-based and vertical-driven information retrieval

PDF

70

Int'l Conf. Artificial Intelligence | ICAI'13 |

Knowledge-based and vertical-driven information retrieval

Gavin Matthews

Vertical Search Works gmatthews@verticalsearchworks.com

Max Petrenko

Vertical Search Works mpetrenko@verticalsearchworks.com

Abstract – The paper introduces the architecture and functionality of the knowledge-based information retrieval technology developed at Vertical Search

Works. A large-scale language-independent ontology is used during indexing, query analysis, and document retrieval as part of a web-scale vertical search engine.

Three specific areas are examined: the knowledge resource, its visualization and editing toolbox, and ways of fine-tuning the ontology to enable intrinsic consistency checks will be discussed; using the knowledge resource for indexing and query analysis purposes, concentrating on how the ontology-driven semantic latching and disambiguation module tackles challenges arising from ambiguous input; tools for evaluating the performance of knowledge-based information retrieval available at Vertical Search

 

Association-Based Identification of Internet Users Interests

PDF

Int'l Conf. Artificial Intelligence | ICAI'13 |

77

Association-Based Identification of Internet Users

Interests

M. Charnine,1 A. Petrov,2 and I. Kuznetsov1

Institute for Informatics Problems, Russian Academy of Sciences, Moscow, Russia

2

Tinkoff Digital, Moscow, Russia

1

Abstract – The method to provide the Internet user with useful information, while using search engines is presented.

Here we mean the systematization of search results according to user’s interests, and also showing advertisement which could be interesting for the user. We introduce concept of

"user profile", consisting of keywords/terms, reflecting user interests. The discovering of such keywords is done by parsing user queries and visited websites. The proposed method uses a tree of categories linked to related websites and to the advertising. From these websites we retrieve primary keywords characterizing categories. The primary keywords are extended with new associated ones (called secondary) which were obtained by the methods of distributive semantics.

 

Automatic Processing and Acoustic Normalization of Vowels

PDF

82

Int'l Conf. Artificial Intelligence | ICAI'13 |

Automatic Processing and Acoustic Normalization of Vowels

Lai, Yi-hsiu

Department of Western Languages & Literature, National University of Kaohsiung

Abstract - This paper addressed the issue of automatic processing of phonetic cues and acoustic normalization of

Hakka vowel production in Mainland China and in Taiwan.

An acoustic study of Hakka vowels, produced by native speakers of Hakka, was conducted. Fifty participants (twenty from Mainland China and thirty from Taiwan) were invited to read a wordlist of Hakka with six vowels embedded in different contexts (i.e. in different tones, in different phrase positions). An automatic approach was then adopted to analyze the collected acoustic data. Acoustic cues of F0, F1,

F2, F3 and F4 were automatically fetched, and these cues were further examined in the normalization procedures. Based on the discussion on the normalization of Hakka vowels, implications for relevant linguistic issues and applications of machine learning to speech and language processing were provided.

 

The Information Structure of the Domain “Intellectual Technologies and Computational Linguistics”: Texts, Corpus, Context

PDF

Int'l Conf. Artificial Intelligence | ICAI'13 |

89

The Information Structure of the Domain

“Intellectual Technologies and Computational Linguistics”:

Texts, Corpus, Context

Elena Yagunova and Anna Savina

St. Petersburg State University, St. Petersburg, Russia

Abstract — The paper presents the study of the terminology, subdomain interaction, information structure of the text and the corpus. The goal of this research is to determine the distribution features of lexis that help to distinguish common and subdomain terminology. Our objective is to identify keyword features as the most informative structural elements, describing the scientific domain of the corpus and the text position within this corpus.

We restricted the task to the study of the following:

• the text as a context for the keyword;

• the corpus as a context for the text.

The scientific domain of the corpus (corpora) varied from “Corpora Linguistics” (monothematic) to “Theory of

Language & Computational Linguistics” (polythematic).

2

Keywords: information structure, domain “intellectual technologies and computational linguistics”, texts, corpus, context

 

Optimal Automated Method for Collaborative Development of University Curricula

PDF

96

Int'l Conf. Artificial Intelligence | ICAI'13 |

Optimal Automated Method for Collaborative

Development of University Curricula

M. Charnine1 and V. Protasov2

Institute for Informatics Problems, Russian Academy of Sciences, Moscow, Russia

2

The Russian Center of Computing for Physics and Technology, Moscow, Russia

1

Abstract – Aspects of curriculum design automation using semantic analysis of Web texts to develop candidate keywords and voting alternatives, as well as methods of optimal aggregation of individual expert decisions are presented. It is shown how semantics analysis of online texts can help with: generating of initial version of the curriculum; generating of candidate keywords and alternative options for improving the curriculum; building hierarchy of categories of the subject area or domain; detecting the change of domain knowledge state and defining structural knowledge dependencies; monitoring terminology and knowledge base of the subject area or domain. It is shown how the techniques of optimal integration of individual expert decisions in selecting of automatically generated alternatives helps to create the highest quality curriculum.

 

Database of Equivalent Verbal Forms in a Russian-French Multivariant Parallel Corpus

PDF

Int'l Conf. Artificial Intelligence | ICAI'13 |

101

Database of Equivalent Verbal Forms in a Russian-French

Multivariant Parallel Corpus

Anna A. Zalizniak1, Dmitri V. Sitchinava2, Sylvain Loiseau3, Mikhail Kruzhkov4 and Igor M. Zatsman4

1

Institute of Linguistics, Institute of Informatics Problems, Russian Academy of Sciences Moscow, Russia

2

Institute of the Russian Language, Russian Academy of Sciences Moscow, Russia

3

Institute Laboratoire LDI (Lexiques, Dictionnaires, Informatique), Université Paris 13, Paris, France

4

Institute of Informatics Problems, Russian Academy of Sciences Moscow, Russia

Abstract - The Russian-French parallel corpus as a part of the Russian National Corpus is being transformed into a multivariant corpus where several translation variants correspond to one and the same original text. On the basis of this corpus, a Database of functionally equivalent lexicogrammatical verbal forms is being created. The main purpose of database creation is to calculate statistical estimates of equivalences between Russian and French verbal forms. The paper discusses information technology for creating the

 

Load more


Details

Print Book
E-Books
Chapters

Format name
PDF
Encrypted
No
Sku
B000000030937
Isbn
9781601322470
File size
868 KB
Printing
Allowed
Copying
Allowed
Read aloud
Allowed
Format name
PDF
Encrypted
No
Printing
Allowed
Copying
Allowed
Read aloud
Allowed
Sku
In metadata
Isbn
In metadata
File size
In metadata