Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Information Analysis Knowledge Mining/Threat Prediction SMU CSE 8394 Organizational Knowledge Knowledge Knowledge we have and don’t know we have it and do not use it Knowledge we have and know we have it and use it Meta-Knowledge Discover!! SMU Knowledge don’t we have and we know that X CSE 8394 Knowledge Discovery Knowledge Discovery Link Analysis Data Mining Unstructured Data SMU CSE 8394 Data Mining • Process of extracting hidden knowledge from raw data – – – – – Targeted marketing Cross-sell opportunities Customer profiling Supply chain? Power, traffic distribution • Why data mining – Huge volumes of data – Limitations of human analysis – Low cost of machine learning (bigger trend that outsourcing!!) SMU CSE 8394 DM Methods • • • • • • Neural Networks Evolutionary Programming Memory Based Reasoning Decision Trees Genetic Algorithms Nonlinear Regression Methods SMU CSE 8394 Graph-Based Link Analysis • The “friends” problem – Understanding the underlying relationships – Prediction of future relationship • Premise – Entities connected through noisy relationships • Question to answer – Given the collaborative graph, who are the top K friends? – Who could be friends in future? SMU CSE 8394 Text Mining with Link Analysis • Reported experiments on – Connection between a terrorist and known figure – Meetings between US and Israel • Approaches to link construction – Co-occurrence links – Semantic links • Quality metrics – Accuracy – Recall SMU CSE 8394 Link Discovery in Battlespace • Information fusion should progress beyond current level 1 (object refinement) • Focus on Continuous Intelligence Preparation of the Battlespace (CIPB) – Develop DA to aid human analyst to determine enemy COA – Leverage technologies such as mobile agent architecture, link discovery and analysis • Sample scenario analysis SMU CSE 8394 Related Message Traffic • Constructing conversations when conventional markers are missing – Look for statistical anomalies – Application of SVD and ICA (Independent Component Analysis) to noun frequency profiles • Application in counter terrorism and fraud SMU CSE 8394