Tiejun Hou

What it does

My system has three purposes:
First level: knowledge base level. I hope one day, everyone can have a knowledge base, and can
store all their books, from their first primary school painting to their last words of life.
Second level: book level. In fact, the main idea of this part is from How To Read a Book, building a
skeleton of a book. This part focuses on extensive reading, the most important task of this level is
to understand the text outside the sentence domain — full text understanding.
The third level of my system is page level, it focuses on intensive reading. The most important part
at this level is inside the sentence.
Why do I call the third level the page level, not sentence level? Because the page concept is the
most important and special concept in the structure of a book, all other concepts of the book are
logic concepts: chapters, sections, sentences. Only the page concept is a physical concept, and
this concept is important in both paper books and E-books. And second level — full text
understanding and the third level — sentence understanding, will encounter and integrate in the
page level. From book level to sentence level, is a deductive process, and from sentence level to
book level, is an inductive process, and the two processes meet in physical concept----page.
I use Gemini in two parts: first, I use Gemini to generate book info; second, I put Gemini interface into my pages, when users are reading, they can communicate with Gemini anytime.