5) Symptom Name Validation: The 'learn' word in the
previous step must be validated manually. Incorrect result is
deleted from the file.
B. Thai herb treatment information extraction
After 'learn' the symptom name, now the process is ready
to extract Thai herb treatment information from webpages. The
overall process is shown in Figure 12.
The process starts from extracting HTML source code
from the webpage, identifYing relevant treatment information,
extracting Thai herb names, extracting part of used, and
extracting symptom name. Some steps are similar to steps in
the previous process; hence, we will explain only some key
steps here.