Sidang proposal english


Published on

Published in: Technology, Education
1 Comment
  • how to save/download?
    Are you sure you want to  Yes  No
    Your message goes here
No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide
  • Latar belakang, Pengenalan ERPGame, Metode Penelitian, WDAG, Semantik Web Service, Agent WS (Multi-Criteria Negotiation)
  • Semantik Web: easier to find, share, agregate and extend informationSemantik WS: easier to discover, use/invoke, compose and monitor.
  • Menggunakan metadata semantik (faktor kunci: representasi metadata)Ceritakan tentang penelitian sebelumnyaSalah satu yang paling banyak digunakan adalah ontology (OWL)Shg, Nyambung ke slide berikutnya tentang wDAG
  • Sidang proposal english

    1. 1. Proposal Thesis<br />Agent-Based Semantic Web Service Using Weighted Directed Acyclic Graph On ERP Game Simulation<br />Oleh:<br />ANANG KUNAEFI (5110 201 008)<br />
    2. 2. 2<br />Agenda<br />Backgrounds<br />Semantic Web Service<br />Weighted Directed Acyclic Graph (wDAG)<br />ERP Game Simulation <br />Research Methodology<br />Summary<br />
    3. 3. 3<br />Latar Belakang<br />The semantic web technology has been widely used for many fields like search engine, content-based application, e-commerce, e-learning, and many others.<br />On the other hand, web service has been widely accepted in the field of business to process daily operation. <br />The needs of semantic web technology to be implemented in the fields of web services in order to promote dynamically integrated application environment.<br />Semantic Web & Semantic WS<br />
    4. 4. Semantic Web dan Semantic WS<br />4<br />Semantics<br />Semantic Web<br />Easier to find, share, aggregate, and extend of content.<br />Semantic Web Service<br />Easier to discover, invoke, compose, and monitor application.<br />Reference:<br />Accessed on August 13, 2011<br />Semantik WS<br />4<br />
    5. 5. Semantik Web Service<br />5<br />Request<br />Service Repository<br />Matchmaking with all WS<br />Discoverer<br />Data Mediator<br />uses<br />uses<br />Composer<br />If: directly usable<br />If: composition needed<br />Communication<br />Conformance<br />Process Mediator<br />Semantic WS can do the following automatically<br />Service Discovery<br />Service Composition<br />Service Enactment & Monitoring<br />Service Negotiation & Contracting<br />uses<br />If: directly compatible<br />Executor<br />If: succesful<br />else: try other WS<br />Reference: University of Innsbruck, Austria (<br />Lecture Material, accessed on August 14, 2011<br />
    6. 6. 6<br />Semantic Web Service Metadata<br />Key factor of semantic web service is how to represent the metadata of web service (Fensel, 2007).<br />Several approaches have been proposed:<br />OWL-S<br />IRS-III<br />WSMO<br />METADATA<br />WEB SERVICE<br />OWL-S<br />
    7. 7. Ontology Web Language for Services (OWL-S)<br />One of the most widely used to represent web service metadata is OWL-S.<br />OWL-S divide service information into :<br />Service Profile<br />Service Grounding<br />Service Model<br />Tools for using OWL-S areProtégé Editor, OWL-S matcher.<br />7<br />Sumber:<br />diakses pada 13 Agustus 2011<br />Keterbatasan OWL-S<br />
    8. 8. Limitation ofOWL-S<br />Usinglogic-based reasoner in the matchmaking process because OWL-S defines “is a” relationship between objects. (Li dan Horrock, 2003). <br />Therefore, in the discovery process, we can’t make preference to a single atribute because all the atribut have the same level of preference.<br />8<br />wDAG<br />
    9. 9. Weighted Directed Acyclic Graph (wDAG)<br />An arc-labeled, arc-weighted DAG is constructed from a 6-tuple (V, E, LV, LE, LW, r) of a set of nodes V, a set of arcs E, a set of nodelabels LV, and a set of arc labels LE, a set of arc weights LW = [0,1], and one element r where r ϵV.<br />wDAG similarity computation is more eficient then weighted tree similarity because wDAG structure is more efficient.<br />This schema can also be used by user/consumer of WS refine the discovery of services by making preferences for some atributes by providing greater weight than other atributes.<br />9<br />ERPGame<br />
    10. 10. ERPGame<br />Enterprise Resource Planning (ERP) Game is learning-by-doing-based games to help the players understanding the concept of ERP (Enterprise Resource Planning).<br />10<br />ERPGame Concept<br />
    11. 11. Konsep ERPGame (1)<br />ERPGame is a unique business simulation technology that enables the simulation of near-real-life ERP business context of corporate information system.<br />It provides the simulation of a market for buyers so that the participants playing the game have a reasonable market that responds just like one in the real world.<br />It automates some of the business functions that are more administrative to make the game a little easier to play so the participants focus on the decision making processes.<br />It provides simulation of the passing time. It compresses time into short but still create the appearance of time evolving so that the impact of the decisions taken vertime can be evaluated.<br />11<br />ERPGame Concept 2<br />
    12. 12. ERPGame Concept (2)<br />ERPGame provides these 3 functions so the game can be played:<br />Provides market for buyers that respond just like in the real world.<br />Provides some business process automation.<br />Provides time simulation.<br />12<br />ERPGame<br />Web service-base ERP System<br />ERP Database<br />ERPGame run on top of <br />Web service-based ERP<br />ERPGame dan Web Service<br />
    13. 13. WS<br />Pembeli/Pasar Virtual<br />ERPGame dan Semantik Web Service<br />Dalam 1 siklus permainan ERPGame diikuti oleh 2 atau lebih player.<br />Tiap player harus menjual produk masing-masing<br />Pembeli/Pasar virtual mencari dan memilih service yang paling menguntungkan.<br />13<br />Team A <br />(role as company A)<br />Mencari service yang sesuai dengan harga yang paling menguntungkan <br />(automatis)<br />WS 1<br />Sell Product A, PriceX<br />WS 2<br />Sell Product A, PriceY<br />Menggunakan<br />Semantik Web Service<br />Berbasis agen<br />Team B<br />(role as company B)<br />Metode Penelitian<br />
    14. 14. Metode Penelitian<br />14<br />Service<br />Service<br />Langkah-langkah dalam pengembangan metode, meliputi:<br />WSDL-S Mining<br />Pembangkitan wDAG<br />Pembuatan wDAG Registry<br />Perhitungan wDAG Similarity<br />Multi-criteria Negotiation<br />WSDL-S Mining<br />WSDL-S Mining<br />(Elgazzar et al., 2010)<br />Pembangkitan WDAG<br />Pembangkitan WDAG<br />(Jin, 2006)<br />wDAG<br />Registry<br />(Nugroho & Sarno, 2011)<br />WDAG<br />Similarity<br />(Jin, 2006)<br />(Mei, 2006) dan <br />(Rao, 2004)<br />Multi-criteria<br />Negotiation<br />Service Agreement<br />WSDL-S Mining<br />
    15. 15. WSDL-S Mining<br />Ekstraksi fitur-fitur dari web service, yaitu:<br />Service Content<br />Service types<br />Messages<br />Ports<br />Service Name<br />Proses ekstraksi terdiri dari<br />Parsing WSDL<br />Tag Removal<br />Stemming<br />Function word removal<br />Content word recognition<br />15<br /><ul><li>Pada penelitian ini akan digunakan WSDL-S sebagai pengganti WSDL.
    16. 16. WSDL-S adalah WSDL ditambah anotasi semantik.</li></ul>Pembangkitan wDAG<br />Referensi: Elgazzar et al., 2010<br />
    17. 17. Pembangkitan wDAG<br />16<br />Service <br />Service <br />Profile<br />Service <br />Grounding<br />Service <br />Model<br />Profile<br />Profile<br />Service<br />Desc<br />Service<br />Desc<br />Grounding<br />Category<br />Desc<br />Category<br />Desc<br />Process<br />Operation<br />Desc<br />Operation<br />Desc<br />Desc Text<br />Desc Text<br />Input Type<br />Precondition<br />Type<br />Category<br />Category<br />Desc Text<br />Output Type<br />Operation<br />Explaination<br />Text<br />WSDL <br />Operation<br />Effect<br />Taxonomy<br />Code<br />CatName<br />Operation<br />Name<br />Type<br />Text<br />Operation<br />Explaination<br />Taxonomy<br />URI<br />Port Type<br />Text<br />Reference<br />Text<br />Text<br />Operation<br />Name<br />Reference<br />Text<br />Reference<br />Reference<br />Text<br />Text<br />Reference<br />Text<br />Hasil dari ekstraksi fitur WSDL-S selanjutnya dimasukkan dalam skema wDAG<br />ReferenceDesc<br />URI<br />Domainmodel<br />Text<br />Text<br />wDAG Similarity<br />Referensi: Jin et al., 2006<br />
    18. 18. wDAG Similarity<br />17<br />Service Operation<br />Service Desc<br />Service Operation<br />Service Desc<br />0.5<br />0.5<br />0.5<br />0.5<br />getInvoice<br />Get Invoice By Distributor<br />getInvoice<br />Get Invoice By Factory<br />Perhitungan kemiripan dari Jin, akan divariasi dengan perhitungan cosine similarity dan wordnet.<br />“Get Invoice By Distributor” dan “Get Invoice By Factory” dengan cosine similarity menghasilkan similarity = 0.336 (bukan 0).<br />Dengan cosine sim “Distributor “ dan “Factory” dianggap memiliki kemiripan sama dengan nol (0). Agar lebih akurat digunakan wordnet untuk mengetahui jarak kemiripan antara kata-kata tersebut. Dengan wordnet ternyata antara “Distributor” dan “Factory” kemiripannya = 0.462.<br />Sehingga gabungan antara cosine dan wordnet menghasilkan similarity yang lebih baik yaitu 0.377.<br />Multi-criteria negotiation<br />
    19. 19. Multi-criteria Negotiation<br />18<br />Agent<br />Behaviour Manager<br />Rule Engine<br />Rule Set<br />Sensor<br />Actuator<br />Web Services<br />Setelah semantik web service menemukan web service yang dicari dengan wDAG Sim, selanjutnya agent negosiator sebagai perantara buyer-seller melakukan negosiasi, untuk menemukan service yang paling menguntungkan. <br />Negosiasi dilakukan atas 3 atribut, yaitu:<br />Harga produk<br />Lokasi<br />Kecepatan layanan<br />Uji coba<br />Referensi: Mei et al., 2006<br />
    20. 20. Uji coba<br />Uji coba terhadap framework yang dibangun meliputi 2 hal, yaitu:<br />Pengujian discovery of service menggunakan precision, recall dan F-Measure (Baeza-Yates, 1999) serta ROC (Receiver Operating Characteristic) Curve.<br />Pengujianapakah service yang terpilihmelaluiprosesmulti-criteria negotiationbenar-benar service yang terbaikdan paling menguntungkan.<br />19<br />
    21. 21. Penutup<br />Tujuan dari penelitian ini adalah untukmendapatkansebuahmetodesemantik web service berbasisagenmenggunakan Weighted Directed Acyclic Graph (wDAG) dalamskema buyer-seller dilingkunganbisnis yang kompetitifpadaERPGame.<br />Manfaat penelitian iniadalahuntukmemudahkanconsumer web service dalammenemukan web service yang paling sesuaidengankebutuhanmerekasecaraotomatisbaikdalamkonteksindividumaupunkonteksbisnisdalamperusahaan.<br />Kontribusi dari penelitian ini adalah:<br />menyediakanskema metadata semantik web service menggunakanweighted directed acyclic graph (wDAG) dengantingkatpenghitungankemiripan yang lebihbaik<br />menyediakan framework untuksemantik web service dalamlingkungan yang kompetitifdenganmulti-criteria negotiationmenggunakan agent web service<br />20<br />
    22. 22. Terima Kasih<br />