The procedure for preparing T-ERD:
　1. Create an entity with the identifier, "xx-NO" or "xx-CODE".
　2. Classify the entities into "event" or "resource", by the criteria of "DATE".
　3. Create relationships between entities, with the following formula:

　□ a ≠ b

　(1) resources-to-resource:

　　　A "combination-table", taisyou-hyo, shall be created, and
　　　the identifiers of the two are posted in a combination-table.

　(2) resource-to-event:

　　　The identifier of the resource shall be posted into the event.

　(3) event-to-event:

　　　If "one-to-one" or "one-to-many",
　　　the identifer (of the preceeding event) shall be posted into the event occurring later in time-sequence.

　　　If "many-to-one" or "many-to-many",
　　　a "mapping-list", taiou-hyo, shall be created, and the identifier of the two shall be posted in a mapping-list.

　□ a = b (recursive)

　　A "recursive" table, "entity. entity. as recurring", shall be created for the "recursive" structure and decomposed into a binary format of "a parent-child" combination.

　Now you can prepare T-ERD!　It's easy, isn't it?

　Well, the T-ERD you've now prepared, how can you prove if it's right (sound and correct)?

　If you cannot prove that it's correct, it may be just an arbitrary work--many system engineers can prepare many diagrams with the same data, and all of the diagrams might be dissimilar.....they cannot be called the products produced through engineering techniques.

　Next time, I'll take a technique, "connotation and extension", to prove that the data clusters you've created should be right.