1
BigDAK@DLMU Since 2021
Name: SATTAR ABDUL
Chinese: 赛塔
Master Student
Major (AI)
Registration 1120248051
2
BigDAK@DLMU Since 2021
Bridging Semantic Maritime Search and Intelligent
Reasoning with Generalized Vectors and LLMs on
PostgreSQL
Topic
3
BigDAK@DLMU Since 2021
Top-tier Journals and Conferences in Your Field
1. Top-tier Journals and Conferences in Your Field
Journals:
• Information Processing & Management
• IEEE Transactions on Knowledge and Data Engineering (TKDE)
• ACM Transactions on Information Systems (TOIS)
• Data Intelligence (MIT & Chinese Academy of Sciences)
• Journal of Web Semantics
Conferences:
• ACL (Association for Computational Linguistics)
• EMNLP (Empirical Methods in NLP)
• ICDE (IEEE International Conference on Data Engineering)
• SIGIR (Special Interest Group on Information Retrieval)
• NeurIPS (Conference on Neural Information Processing Systems)
4
BigDAK@DLMU Since 2021
Why Papers Published in These Journals/Conferences Have High Impact
Relevance to Industry and Academia:
• They align with cutting-edge trends in semantic search, vector databases, and LLMs.
Cross-disciplinary Reach:
• Many papers connect maritime analytics, data engineering, NLP, and AI.
Citation Power:
• Published works often define foundational benchmarks and are widely cited.
Real-world Impact:
• Focused on deployable solutions in smart transportation, maritime safety, and knowledge-based
systems.
5
BigDAK@DLMU Since 2021
Your Research Title, Problem, and Hypothesis
• Research Problem: Maritime data is vast, heterogeneous structured and
unstructured, and often lacks semantic search capabilities. Traditional keyword-
based systems fail to retrieve contextually relevant insights, which impairs
navigation safety, risk analysis, and maritime event prediction.
• Hypothesis: Integrating vector databases (e.g., pgvector on PostgreSQL) with
domain-tuned LLM embeddings (e.g., BGE, MiniLM) can enable scalable,
semantically meaningful search and reasoning over maritime data, improving
safety, efficiency, and intelligence in maritime operations.
6
BigDAK@DLMU Since 2021
Reasons for Choosing This Research Topic
Practical Importance
• Maritime traffic management, port optimization, and risk analysis require intelligent data systems
Technological Trend
• Combining pgvector, LLMs, and open-source vector search reflects a modern, scalable architecture being adopted globally.
Academic Gaps
• Few papers have applied LLM-driven semantic search to large-scale maritime data.
Supervisory Alignment
• supervisor’s recommendation to explore Chinese models (e.g., DeepSeek, Gimi, ZhipuAI) aligns with this direction.
Career Goals:
• This topic strengthens your expertise in AI, vector search, and applied NLP all vital for research or data scientist roles.
7
BigDAK@DLMU Since 2021
Thank You
Any Questions

Semantic Martime search under knowledge graphsPPT.pdf

  • 1.
    1 BigDAK@DLMU Since 2021 Name:SATTAR ABDUL Chinese: 赛塔 Master Student Major (AI) Registration 1120248051
  • 2.
    2 BigDAK@DLMU Since 2021 BridgingSemantic Maritime Search and Intelligent Reasoning with Generalized Vectors and LLMs on PostgreSQL Topic
  • 3.
    3 BigDAK@DLMU Since 2021 Top-tierJournals and Conferences in Your Field 1. Top-tier Journals and Conferences in Your Field Journals: • Information Processing & Management • IEEE Transactions on Knowledge and Data Engineering (TKDE) • ACM Transactions on Information Systems (TOIS) • Data Intelligence (MIT & Chinese Academy of Sciences) • Journal of Web Semantics Conferences: • ACL (Association for Computational Linguistics) • EMNLP (Empirical Methods in NLP) • ICDE (IEEE International Conference on Data Engineering) • SIGIR (Special Interest Group on Information Retrieval) • NeurIPS (Conference on Neural Information Processing Systems)
  • 4.
    4 BigDAK@DLMU Since 2021 WhyPapers Published in These Journals/Conferences Have High Impact Relevance to Industry and Academia: • They align with cutting-edge trends in semantic search, vector databases, and LLMs. Cross-disciplinary Reach: • Many papers connect maritime analytics, data engineering, NLP, and AI. Citation Power: • Published works often define foundational benchmarks and are widely cited. Real-world Impact: • Focused on deployable solutions in smart transportation, maritime safety, and knowledge-based systems.
  • 5.
    5 BigDAK@DLMU Since 2021 YourResearch Title, Problem, and Hypothesis • Research Problem: Maritime data is vast, heterogeneous structured and unstructured, and often lacks semantic search capabilities. Traditional keyword- based systems fail to retrieve contextually relevant insights, which impairs navigation safety, risk analysis, and maritime event prediction. • Hypothesis: Integrating vector databases (e.g., pgvector on PostgreSQL) with domain-tuned LLM embeddings (e.g., BGE, MiniLM) can enable scalable, semantically meaningful search and reasoning over maritime data, improving safety, efficiency, and intelligence in maritime operations.
  • 6.
    6 BigDAK@DLMU Since 2021 Reasonsfor Choosing This Research Topic Practical Importance • Maritime traffic management, port optimization, and risk analysis require intelligent data systems Technological Trend • Combining pgvector, LLMs, and open-source vector search reflects a modern, scalable architecture being adopted globally. Academic Gaps • Few papers have applied LLM-driven semantic search to large-scale maritime data. Supervisory Alignment • supervisor’s recommendation to explore Chinese models (e.g., DeepSeek, Gimi, ZhipuAI) aligns with this direction. Career Goals: • This topic strengthens your expertise in AI, vector search, and applied NLP all vital for research or data scientist roles.
  • 7.