Search and Find with IAI’s FindAUT Tool


Intelligent and Efficient Content Extraction

Efficient information gathering from an ever growing sea of  information is a key competitive factor in the era of information society. Providing easily accessible high quality information for customers and employees contributes to customer satisfaction and efficiency of workflows within the company. Our software suite FindAUT automatically indexes large stocks of documentation thus assigning key words through which a focused search becomes available. The techniques based on a combination of linguistic and statistical methods provides unique quality. 

FindAUT integrates AUTINDEX into the Lucene based search engine SOLr thus providing a search facility far beyond full text search which is especially relevant for languages with rich inflection.

FindAUT uses background knowledge by integrating thesauri into the extraction and the search strategy. Statistical weighting mechanisms further enhance both extraction and search.

Quality through Linguistic Intelligence

Linguistic analysis is based on a morphological analysis with:

  • Lemmatisation
  • Compound analysis
  • Checking of orthographioc correctness

A sophisticated grammatical analysis component contributes to a disambiguation of ambiguous word forms.

The integration of a thesaurus expands search space systematically by

  • Morpho syntactic transformations
  • By ‘synonymisation’ of components of compounds

To give an illustration: The search term 'Netzstabilität' may also find documents in which the same concept may be referred to by 'Stabilität des Netzes'. (An English example: Cost reduction vs. reduction of costs). The second point may be illustrated by Emissionsreduktion vs. Emissionsminderung. (An English example: reduction of costs vs. saving of costs). 

Easy Accessibility for User Convenience

The calculation of a relevance factor for each key term results in so called word clouds which are meant to support the user in a facetted search. Word clouds may be used as a filter to further constrain the search results. Word clouds may also be used to exhibit the most prominent key terms that best characterise the contents of a document in relation to the whole subject area.