Does e-FAQ use case based reasoning, or CBR?.
e-FAQ does not use case based reasoning or backward-chaining logic methods. e-FAQ uses linguistic-based indexing and matching. From a linguistic standpoint, e-FAQ uses morphological stemming, a semantic network for synonyms, and other assorted linguistic algorithms. e-FAQ also performs an informational and statistical analysis of the FAQ database to determine important words. E-FAQ also uses morphological stemming and a semantic network for synonmyms. In addition, it uses an innovative “dynamic-thresholding” technique for evaluating match quality. Finally, e-FAQ uses machine learning techniques to incorporate both wording and entry-confidence feedback.
