<?xml version="1.0" standalone="yes"?>
<Paper uid="N03-4002">
  <Title>Alias-i ThreatTrackers(TM)</Title>
  <Section position="4" start_page="0" end_page="0" type="metho">
    <SectionTitle>
2. Cognitive Indexing(TM)
</SectionTitle>
    <Paragraph position="0"> Cognitive Indexing organizes text data feeds into mentally natural units like people, events, things, organizations and relationships. For example, the Cognitive Indexing representation for an Entity like Osama bin Laden contains pointers to all sentences in the data feed that mention him Fig 4, including spelling variation (i.e. Usama) and pronouns which refer to him. In future versions of the system, the Cognitive Indexing representation of Osama will include audio mentions of him as well as database entries for him. In addition to coreference, Cognitive Indexing supports standard key word lookup, relationship detection between entities and analyst authored classes of Entities.</Paragraph>
  </Section>
  <Section position="5" start_page="0" end_page="0" type="metho">
    <SectionTitle>
3. ThreatTracker Interface
</SectionTitle>
    <Paragraph position="0"> The ThreatTracker interface provides complete analyst control over what is in a ThreatTracker, and provides multiple views into the collected data. Included are sentence excerpt summaries, tables of relationships between Entities and easy access to original source documents. There is also a redundant information filter on documents or sentences. ThreatTrackers allow analysts to create new categories of Entities like &amp;quot;Chemical Weapons Precursors &amp;quot; for use inside ThreatTrackers. Once created, ThreatTrackers serve as a form of information triage against the data feed.</Paragraph>
  </Section>
class="xml-element"></Paper>