In the course of many years, Inquera gained unique experience in product
data and product
data quality.
As a natural development, Inquera decided to co-defy this knowledge into an automatic
software system.
Meaning, developing methods and technologies with the most efficiency and reliability in dealing with the product data quality great challenge.
There are a few popular methods of product
data handling:
1. Manual handling by internal staff or by external service providers.
2. Writing "business rules" and lexicons and applying matching algorithms.
3. Applying semantic rules and lexicons.
The main problems with these approaches are:
1. Manual handling is inefficient, time consuming, expensive, not scalable and cannot
produce high quality results.
2. Business rules might be effective on structured data in a specific language,
but the
combination of unstructured data and multilingual product descriptions, means, in effect,
writing a specific business rule for almost each product
description.
Writing complex business rules requires subject domain experts
and such experts are not
necessarily trained in writing business rules.
3. Semantic engines are in a sense rule-engines based on semantic rules, the problem is
that unstructured data has no semantic.
After understanding the disadvantages of the existing common approaches it became clear to Inquera that the only appropriate technology direction to adapt
is AI (artificial intelligence) machine learning (self learning) approach.
Inquera's technology is based on the ability of systems to learn from examples provided by subject domain experts.
In effect Inquera's systems request the subject domain experts to perform a few manual activities such as: classification, or extraction. These activities become
the knowledgebase which embodies the accumulated knowledge of this domain.
Inquera's technology is completely indifferent to: language, dialect, abbreviations
and structure.
Inquera's technology is efficient, scalable, reliable, and cost effective, and preserves the ongoing engineering & organizational knowledge.
|
