Home | IEBuilder Toolkit | Experience | About Us | Contact Us | Products and Pricing | Customers

infoextract_header.medium.cropped.tm.jpg

Experience

The IEBuilder ™ Toolkit is based on decades of experience in developing successful natural language processing (NLP) applications. This experience teaches many hard-won lessons:

  • Data Analysis Pays Big Dividends : analysis of data representative of the intended task and users reveals beneficial and sometimes unexpected insights
  • Expect to be Wrong : initial ideas of how language works are often defective and sometimes way off the mark
  • Measure Twice, Cut Once : evaluation -- testing performance objectively against standard data samples -- exposes the issues that matter for product development and provides the means to rationalize and prioritize them
  • There's More Than One Way to Do It : there's often no single best way to implement a language processing task -- and frequently a combination of techniques works even better
  • Developing for the "Average" User Means You're Wrong on Average : if you have no other alternative, you should develop for an "average" user, but analysis during development rarely anticipates how a specific individual works


We Did It, So You Won't Have To

These lessons have guided the design and feature set of the IEBuilder SDK. IEBuilder has a unique combination and tightly integrated suite of tools for developing natural language (NLP) applications:

  • multiple NLP technologies and approaches
  • data exploration and discovery tightly integrated into product development and maintenance
  • easy deployment of developed systems
  • built-in adaptation to new users, new usage patterns, and data not anticipated during development
  • pre-built language components
  • slipstreaming your legacy or other third-party language processing technologies

Using the IEBuilder ™ Toolkit
  • Multi-Approach Analysis Engine : use built-in pattern-based, machine-learning, and rule-based NLP technologies

  • Integrated Development and Exploration Environment : Discover language patterns in your data efficiently and incorporate this knowledge in your applications.

  • Corpus Management : Select, sample, annotate, view, categorize, and edit large document collections

  • Annotation : Label and validate data in your document collections using efficient automated methods

  • Run-Time Environment : Convert a development system into a deployed system seamlessly and efficiently

  • Feedback in Development and Run-time Environments : Use feedback to improve product quality and user experience

  • Pre-Built English Language Processing Tools : High-quality and modifiable English language processing tools for tokenization, sentence-boundary detection, inflectional and derivational morphology, part-of-speech analysis, phrase analysis, parsing, gazetteers, ontologies, and spelling correction

  • Customizable Workflows : Construct application and language processing workflows using built-in high-level programming language and rules-based capabilities

  • Use What You've Already Done : Incorporate already existing or third-party language processing tools

  • Roll Your Own : Prototype or make production-ready natural language processing and knowledge management tools using IEBuilder's built-in high-level programming language and tools

  • Portable : Written in platform-independent Java

  • Standards : Uses Unicode character sets and built-in Unstructured Information Management Architecture (UIMA) technology

For more information about IEBuilder, click here.

Information Extraction Systems, Inc. * 20 East Quinobequin Road * Waban MA * USA * 02468 Phone: (617) 244-5068 Fax: (617) 244-5068

The InfoExtract mark and logo are either trademarks or registered trademarks owned by Information Extraction Systems, Inc. All other trademarks are owned by their respective owners.

Copyright 2007, Information Extraction Systems, Inc. All rights reserved.