I am currently working on something similar in my final thesis - kanji categorized into around 75
mostly semantic groups, however my system makes use of fundamental components and on readings as well. I will show you some "samples", as the thesis has not been published, yet. For example, the first group representing "numbers"
without having the uniform fundamental component looks like this:
a) 一二仁三四五六七八九十百千万億憶兆眺挑跳
一二仁三四五六七八九十百千万億兆 = chief characters ... share the same semantic concept ("numbers")
仁憶眺挑跳 = visual-reading companion characters ... not related to the group semantically, but they share the same component as one of the chief characters + at least one same on (Sinojapanese) reading
- most of the groups actually
share the same fundamental component (radical) as well:
b) 魚鮎鯉鯛鮭鮫鯨鮮漁
魚鮎鯉鯛鮭鮫鯨 = semantic chief characters ... share the same semantic concept ("fishes")
鮮漁 = visual chief characters ... does not share the semantic concept, but contains the same fundamental component
(魚)
Let me know in case you are interested in my list. My mail is:
kandrac.patrick@gmail.com