RESEARCH Projects, Year 1

THEME 3.1 Projects on Collecting and Building Knowledge from Data

• Research Project 1:
  Extracting Information from Diverse Media Sources
  Goals: collecting information from various sources and media, and extracting into common framework.
  Tasks starting in year 1:
  1. Extracting relations from text
  2. Extracting events, opinions, and other information from text
  3. Extracting pertinent information from text and images

• Research Project 2:
  Identifying Information Needs, Distillation, and Filtering
  Goals: Ranking and distilling information; initial integration into centralized information structures and repositories.
  Tasks starting in year 1:
  1. Methods to perform data distillation
  2. Dataless categorization
  3. Data integration and holistic search

THEME 3.2. Projects on Enriching Knowledge using Inference

• Research Project 3:
  Storing and Managing Large Amounts of Information
  Goals: Construction of overall analytics framework and supporting databases and knowledge bases.
  Tasks starting in year 1:
  1. Methods to construct and maintain central knowledge repository (large framework that incorporates diverse kinds of facts and knowledge from multiple media, including integrating ontology)
  2. Topic analysis as support for knowledge integration

• Research Project 4:
  Inference and Textual Entailment
  Goals: Deployment of inference engine and reasoning in order to find knowledge gaps and inconsistencies, to locate anomalies, and to identify interesting information to propose to analyst.
  Tasks starting in year 1:
  1. Applying PowerLoom inference engine to find knowledge gaps and inconsistencies
  2. Learning and inference for textual entailments

• Research Project 5:
  Hypothesis Formation and Anomaly Detection
  Goals: Form and evaluate hypotheses from text and other information. Induce hypotheses, propose them to analyst, and interact in order to rate, refine, and apply them during knowledge discovery and analysis.
  Tasks starting in year 1:
  1. Forming and evaluating hypotheses from text
  2. Anomaly detection in time-varying multigraphs
  3. Understanding the Interplay between Reporting, Information Flow, and Observed Disease Dynamics

THEME 3.3 Projects on Applying and Sharing Knowledge

• Research Project 6:
  Privacy, Anonymization, Trustworthiness and Sharing
  Goals: Develop theory and software tools focusing on questions of privacy, anonymization, trustworthiness, and sharing.
  Tasks starting in year 1:
  1. Trustworthiness of online documents and data sources
  2. Anonymization of sensitive information
  3. Differential privacy

• Research Project 7:
  Tools for Cybersecurity
  Goals: Create Analytics tools that support cybersecurity research and that integrate cybersecurity tools into CCIC and other research projects.
  Tasks starting in year 1:
  1. Machine learning methods for cybersecurity
  2. Hypothesis-driven dynamic data analysis tools for cybersecurity: modeling and detection of abnormal behavior in network traffic
  3. Cybersecurity tools: provenance tracking, trust
  4. Secure mashup framework for dynamic integration and analysis

THEME 3.4 Projects Developing Foundations for Homeland Security Tools and Applications and for Collaborating with other DHS Centers of Excellence and with the Visual Analytics Half of CCI

• Research Project 8:
  Foundations for Homeland Security Tools
  Goals: Develop foundational theory and methodologies of Analytics as basis for other projects and applications. Numerous quite different alternatives to choose from here; many kept for later years.
  Tasks starting in year 1:
  1. Exploiting games: Virtual worlds
  2. Optimal sampling
  3. Sequential inspection methods (with applications to container inspection, email/website testing, etc.)
  4. Learning and probabilistic modeling
  5. Massive graphs
  6. Methodology of evaluating Analytics

• Research Project 9:
  CCICADA-COE collaborations
  Goals: Work closely with selected members of DHS COEs to identify and solve pressing problems of general relevance.
  Tasks starting in year 1
  1. RU-FAZD collaboration: Biosurveillance
  2. CCICADA-START collaboration: Adding Analytics to Global Terrorism Database
  3. RU-CREATE collaboration: Port Resiliency - Risk Assessment

• Research Project 10:
  Data Sciences-Visual Analytics Collaborations
  Goals: Research and develop frameworks in which discrete analytics theories and modules can be tested and illustrated. Work with Visual Analytics team members on joint projects.
  Tasks starting in year 1
  1. Crime Analysis, built on Port Authority of NY/NJ project,includes data analysis and VA
  2. Data sciences-visual analytics collaborations, Geojunction+GeoSemble
  3. Data Streams and Network Data (Project in year 2 to be named: Incorporating data analytics tools into Jigsaw)
  4. Port Security - See Project 8.3 Sequential Inspection Methods.
  5. Re-engineer PNNL Wall.