Presentation is loading. Please wait.

Presentation is loading. Please wait.

Dynamic Matchmaking between Messages and Services in Multi-Agent Systems Muhammed Al-Muhammed May 3, 2004 Support in part by NSF.

Similar presentations


Presentation on theme: "Dynamic Matchmaking between Messages and Services in Multi-Agent Systems Muhammed Al-Muhammed May 3, 2004 Support in part by NSF."— Presentation transcript:

1

2 Dynamic Matchmaking between Messages and Services in Multi-Agent Systems Muhammed Al-Muhammed May 3, 2004 Support in part by NSF

3 2 Motivations Agents cooperate to achieve goals Cooperation needs communication Communication possible if agents: –share ontologies, –speak the same language, –pre-agree on a message format.

4 3 The Problem Requiring these assumptions precludes agents from interoperating on the fly “T he holy grail of semantic integration in architectures” is to “allow two agents to generate needed mappings between them on the fly without a priori agreement and without them having built-in knowledge of any common ontology.” [Uschold 02] Agents must: 1- share ontologies, 2- speak the same language, 3- pre-agree on message format.

5 4 Solution Eliminate all assumptions Agents must: 2- speak the same language, 3- pre-agree on message format. - Dynamically capturing a message’s semantics - Matching a message with a service - Translating (developing mutual understanding) This requires: 1- share ontologies,

6 5 Matchmaking System Mapping (MMS ) Message Handling MatchMaking System Service Analysis Translation Response Handling Message-Service Matching Global Domain Ontology An Agent Translation Repository Services (Agent- Independent Representation) Local Ontology Services

7 6 Global Ontology Creation Mapping (MMS ) Message Handling MatchMaking System Service Analysis Translation Response Handling Message-Service Matching Global Domain Ontology Translation Repository Services (Agent- Independent Representation) Concept Recognizers: ProcessorType: (Processor)(Type|Class) ProcessorSpeed: (Processor)(Speed)|(Processor)(Clock)(Speed) … Unit of Measurement Recognizers: ProcessorSpeed Unit: (GHz|MHz) … An Agent

8 7 Local-Global Mappings (Initialization) Mapping (MMS ) Message Handling MatchMaking System Service Analysis Translation Response Handling Message-Service Matching Global Domain Ontology An Agent … double ProcessorClockSpeed; //GHz String ProcessorClass; … Translation Repository Services (Agent- Independent Representation ) Concepts: (Local, Global) ------------------------------------ (ProcessorClockSpeed, ProcessorSpeed) (ProcessorClass, ProcessorType) … Units: ProcessorSpeed: GHz

9 8 Service Analysis (Initialization) Mapping (MMS ) Message Handling MatchMaking System Service Analysis Translation Response Handling Message-Service Matching Global Domain Ontology Translation Repository Services (Agent- Independent Representation ) An Agent … public PcInfo getPcInfo (double RAM){…} public int getPrice (String ProcessorClass, double ProcessorClockSpeed ) {…} //output: Price public int getAmt(String Processor ) {…} //type definition class PcInfo {String ProcessorClockSpeed; //GHz String ProcessorClass; …;} An Agent … public PcInfo getPcInfo (double RAM){…} public int getPrice (String ProcessorClass, double ProcessorClockSpeed ) {…} //output: Price public int getAmt(String Processor ) {…} //type definition class PcInfo {String ProcessorClockSpeed; //GHz String ProcessorClass; …;}

10 9 Requests Rewriting (Initialization) Mapping (MMS ) Message Handling MatchMaking System Service Analysis Translation Response Handling Message-Service Matching Global Domain Ontology Translation Repository Services (Agent- Independent Representation ) An Agent … String ProcessorClockSpeed; //GHz String ProcessorClass; double Price; //US$ Price = getPrice(“ProcessorClockSpeed = 2.6 GHz”, “ProcessorClass = Pentium 4”); MMS.sendString (“ProcessorClockSpeed = 2.6 GHz” ); MMS.sendString,(“ProcessorClass = Pentium 4”); double Price = MMS.sendDouble(“getPrice”);

11 10 Matchmaking System (Operation) Agent 1 Agent 2 Mapping Translation Repository Service Analysis Translation Response Handling Mapping MMS Translation Repository Service Analysis Translation Response Handling Services (Agent- Independent Representation) Services (Agent- Independent Representation) KQML Global Ontology Global Ontology Message Handling Message Handling Message-Service Matching Message-Service Matching String ProcessorClockSpeed; //GHz String ProcessorClass; double Price; //US$ MMS.sendString(“ProcessorClockSpeed = 2.6 GHz”); MMS.sendString(“ProcessorClass = Pentium 4”); Price = MMS.sendDouble(“getPrice”); Price = $1000 MMS ?

12 11 Test Cases Real-World Test Cases –Computer Shopping –Book Shopping –Meeting Scheduling Agents Coded w.r.t. –Each web site (for shopping applications) –Each individual’s worksheet (for scheduling)

13 12 Agent Creation (Concepts & Units) …; ProcessorType; ProcessorSpeed; //GHz …;

14 13 Agent Creation (Services) ReturnType? Name?( Type? InstalledMemory) ReturnInformation? class PcInfo { …; String ProcessorClass; String ProcessorSpeed; //GHz …; } getPcInfo String

15 14 Total number of concepts in agents’ code104 MMS-produced mapping pairs94 Correct mapping pairs91 Recall [Recall = (# of correctly recognized items) / (total # of items that should have been recognized)] 91/104 = 88% Precision [Precision = (# of correctly recognized items) / (total # of recognized items)] 91/94 = 97% Results (Computer Shopping, 9 Agents) Concept RecognitionUnit RecognitionData Format RecognitionConcept Recognition Tested Processes

16 15 Units Currencies: US$, GBP, EUR Number of instances in agents’ code: 9 Capacity and speed: GB, MB, GHz, MHz Number of instances in agents’ code: 23 Total 32 MMS-recognized units 34 Correct units 32 Recall 32/32 = 100% Precision 32/34 = 94% Results (Computer Shopping) Concept RecognitionUnit RecognitionData Format RecognitionUnit Recognition Tested Processes

17 16 No data format of interest Results (Computer Shopping) Concept RecognitionUnit RecognitionData Format Recognition Tested Processes

18 17 Total number of concepts in agents’ code27 MMS-produced mapping pairs25 Correct mapping pairs25 Recall25/27 = 93% Precision25/25 = 100% Results (Book Shopping, 4 Agents) Concept RecognitionUnit RecognitionData Format RecognitionConcept Recognition Tested Processes

19 18 Units Currencies: US$, EURNumber of instances in agents’ code: 4 -- Total 4 MMS-recognized units 4 Correct units 4 Recall 4/4 = 100% Precision 4/4= 100% Results (Book Shopping) Concept RecognitionUnit RecognitionData Format RecognitionUnit Recognition Tested Processes

20 19 Data format Different date formats: 3Number of instances in agents’ code: 4 -- Total 4 MMS-recognized data formats 4 Correct data format 4 Recall 4/4 = 100% Precision 4/4= 100% Results (Book Shopping) Concept RecognitionUnit RecognitionData Format Recognition Tested Processes

21 20 Total number of concepts in agents’ code28 MMS-produced mapping pairs22 Correct mapping pairs22 Recall22/28 = 79% Precision22/22 = 100% Results (Meeting Scheduling, 4 Agents) Concept RecognitionUnit RecognitionData Format RecognitionConcept Recognition Tested Processes

22 21 No units of interest Results (Meeting Scheduling) Concept RecognitionUnit RecognitionData Format Recognition Tested Processes Unit Recognition

23 22 Data format Different date formats: 4Number of instances in agents’ code: 4 Different time formats: 1Number of instances in agents’ code: 4 Total 8 MMS-recognized data formats 8 Correct data format 8 Recall 8/8 = 100% Precision 8/8= 100% Results (Meeting Scheduling) Concept RecognitionUnit RecognitionData Format Recognition Tested Processes

24 23 Contributions Built an MMS that lets agents communicate with no need to –Share ontologies –Use a common language –Pre-agree on a message format Tested the MMS on three applications –Concept mappings (~90% accurate) –Mappings for units and data formats (~98% accurate)

25 24 Future Work Generalize the recognizers and adding some reasoning rules Extend the matchmaking capability to cover partial matching Handle all types of knowledge sharing among agents.


Download ppt "Dynamic Matchmaking between Messages and Services in Multi-Agent Systems Muhammed Al-Muhammed May 3, 2004 Support in part by NSF."

Similar presentations


Ads by Google