master thesis
play

Master Thesis Topic: Tool-chain architecture for data-intensive - PowerPoint PPT Presentation

Master Thesis Topic: Tool-chain architecture for data-intensive computations Aim & Intro Tool integration using Linked data approach This distributed approach is Good Performance Validation Complicated queries


  1. Master Thesis Topic: Tool-chain architecture for data-intensive computations

  2. Aim & Intro ● Tool integration using Linked data approach ● This distributed approach is Good ○ Performance Validation ○ Complicated queries ● Performance enhancement using caching ○ Caching at tool level ○ Caching at tool-chain level

  3. OSLC - Industrial tool integration initiatives ● OSLC standard is a promising solution for tool and data integration.

  4. Tool Chain Architecture Based on OSLC

  5. Adaptor Model - Sessam Tool ● Starting with Resource: ECU Component

  6. Adaptor Model - MSC Drawer ● The ECU is URL pointing to Sesamm

  7. Data Intensive Queries ● Simple Integration Scenario Query ○ Give me all the MSC’s that are relevant for a specific ECU Component ● More Complex Scenario ○ Give me all the User Functions where ECU Component x, is in at-least one of it’s MSC

  8. Tool Chain Architecture ● TRS Provider

  9. Tasks Completed ● Literature Review ○ Linked data & RDF ○ OSLC Domain Specification ● Architecture Implementation ○ IBM tools LQE and RELM Installation Setup ○ Initial Sesamm Adaptor Model ○ Initial MSC Adaptor Model

  10. Tasks Completed .. ● OSLC adaptors Resources ○ Sesamm Tool ● Implement TRS Provider ○ Sesamm Tool BaseLog & ChangeLog ○

  11. Demo ● OSLC adaptor for Sesamm Tool ● TRS Provider For Sesamm Tool ● LQE query capability for Bugzilla & TRS Reference Application ● Challenges & results till now

  12. Future Goals ● Implement ChangeLog for SesammTool ● Implement MSC OSLC adaptor & TRS provider ● Evaluate the Architecture

  13. Future Goals Cont.. ● Investigate if RELM is a Dead End in tool chain? ○ Answer: NO, LifeCycle Query Engine. ● Introduce Caching in Architecture

Recommend


More recommend