A Quick DialogueMaster™ overview

This Site is not supposed to be an advertising or promotional site for the DialogueMaster™, but still I want you to give a quick overview of what the full blown version offers.

• [Det¹] Resolve IP-Adresse to country (helps to disambiguate languages)
• [Heu²] Languagedetection
• [Det] Codepage-Detection
• Usage of the IFilter API (Windows-Search Server) to process different types of files and attachments (PDF, Office etc.)
• Text Splitter
  • Detect headers, Paragraphs, Foots, Disclaims, ads etc.
• [Det] Sentence Breaker(Split long texts into single sentences)
• [Heu} PartOfSpeech-Tagger
• Annotations
  • [Det] Kompostia-splitter (detect and split german compund words)
  • [Det] Word
    • Lemma
    • Stem
    • Detailed wordform information
  • [Det] NamedEntity from TopicMaps
    • Firstname (incl. sex)
    • Cities
    • Countries
    • Companies
  • [Det] Standard Information
    • Numbers
    • Dates
    • Times
    • Timespans
  • [Det] "Codes"
    • Creditcard numbers etc.
  • [Det] Complex Information
    • Addresses
• [Det] Sentence Parser/Chunker
• Classifier
  • [Heu] META (Combined weighted meta classifier)
  • [Det] Rulebased (Regex and custom)
  • [Heu] Metis
  • [Heu] N-Gramm
  • [Heu] Text-Similarity (with linguistic weighting)
  • [Heu/Beta] Lingusitische Templates
  • [Heu/Beta] Image similarity
• Graph based TopicMaps (TMAPI Implementation)
  • Weighted Assoziations
  • Topics are wrapped into a pre-Compilied ontologie (implizitly created from the data)
    • vorhandende TopicMaps für
      • Continents and Countries
      • Germany (NUTS[1-3], Cities)
      • Austria (NUTS[1-3], Cities)
      • Swiss (NUTS[1-3], Cities)
      • Firstname (incl. sex and typical occurences by country)
      • some examples for companies and persons
• Clusterer
• Workflow-Engine
  • Based on .NET Windows Workflow Foundation (WWF) and extended with many custom Activities:
    • File Handling (Move, Copy etc)
    • IMAP
    • SMTP
    • POP3
    • E-Mail Creation
    • Template Engine (based on Apache NVelocity)
    • Classification
    • Information Extraction 
    • Microsoft CRM
    • Perle (Siemens customer interface)
    • CA Unicenter ServiceDesk
• Admin GUI (optimized for learn data management)
  • Sepcial learn text editor
  • Clusterer
  • TopicMap Explorer
  • Index Explorer
  • Regel Editor (Regex)
  • - Meta-Regel Editor (RuleSet Classifier)
• Other
  • Entirely written in C#/.NET
  • runs as Windows Service, Console und rich UI
  • Designed for high paralleisation (Multithreaded)
  • Performance optimized
  • 64Bit tested
  • Uses all avalilabe memory (z.B. für Caches der Wordformen, Indizes, Topics etc.)
  • supports windows performance counters
  • Remote API access via SOAP and Remoting (NamedPipes and TCP) 
    

I plan to create WebsSrvices more or less from top to bottom in this feature list. Due to the complexity of NLP thw WebServices will only offer limited support of the features.

And finally some Screenshots of the GUI:

 
Learn Data Editor



Classification


RuleSet  Editor


Workflow Editor 

¹Deterministic
²Heuristic