Knowledge Discovery and Knowledge Transfer

Abstract

Reasoning-based methods have recently become popular since they overcome the knowledge-acquisition bottleneck during volume production and they can automatically generate an intelligent diagnostic system from existing resources [1, 2]. However, knowledge acquisition is a major problem for a reasoning-based method at the initial product ramp-up stage. Machine learning-based reasoning requires an adequate database for training the reasoning engine, and such a database becomes available much later in the product cycle. In this chapter, we propose a knowledge-discovery method and a knowledge-transfer method for facilitating board-level functional fault diagnosis at the initial product ramp-up stage. The proposed methods help address the knowledge gap between test design stage and volume production. First, an analysis technique based on topic model is used to discover knowledge from syndromes, which can be used for training a diagnosis engine. Second, knowledge from diagnosis engines used for earlier-generation products can be automatically transferred through root-cause mapping and syndrome mapping based on keywords and board-structure similarities.