Tantangan Penterjemahan Mesin

Salah satu tantangan terbesar dalam Machine Translation adalah menterjemahkan joke. Sebagai contoh, cobalah menterjemahkan dua lelucon berikut ini ke dalam bahasa Inggris.

I.
Tahu dan mengerti itu beda. Sama seperti perbedaan antara tahu dan tempe.

II.
T: Bebek. Apa yang enak dari bebek?
J: B-nya. Coba semua huruf B dari Bebek dihapus, jadinya kan nggak
   enak.

Seeing the brain hear reveals surprises about how sound is processed

Speech processing mungkin bisa belajar dari perkembangan baru ini. Bahwa pemrosesan suara yang kita 'dengar' di otak tidak seperti yang dibayangkan sebelumnya.

Seeing the brain hear reveals surprises about how sound is processed: "Neurobiologists have discovered information about how the brain processes sound that challenges previous understandings of the auditory cortex. Their research shows our brains are a lot more chaotic than previously thought, and that this might be a good thing."

Search Engine with Semantic Capabilities

Suppose I need to find a wonderful tool to read my electronic books. Strictly speaking, I have a big collection of e-books in different formats. Most of them are in PDF, some others in PostScript, while there are several e-books in DJVU format. A small portion of my e-library contains files in Microsoft Word *.doc format.

I love reading those books. So I need a great software that can read these formats, not only PDF like Adobe Acrobat. I need one single software that can read PDFs, PSes, DJVUs, DOCs, TXTs, . . . As for another criteria among others, is this software should have a feature that allow me to create bookmarks in all of my e-books. And I also need that the next time I open one particular e-book, I just continue reading from the last page I read. I don't like the idea of finding which page the last time I read an e-book, scrolling down through pages from page one to the last page I read is out of question.

I don't have any idea of the existence of this kind of software. What can I do? If we go to a brainstorming session, here are some options:

I can build that software myself. Maybe I build it from scratch, or using some existing libraries I can find in the internet. The problem with this option is it will take some time for me before I can enjoying reading my e-books.

Another option I can take is I can ask Mr. Google to help me finding such a software. But the problem is what phrases should I type into Google's textbox? Any idea? Maybe you can come up with

"e-book reader", or
"e-book reader with bookmark capabilities", or
"reader software", or
several other phrases you can come up with.

How should you include information to your query that you need such specific features from the software you need to find? At this time all I can think of is entering this long query to Google:

"e-book reader pdf ps djvu bookmark capability windows xp free"

What are these all about?

Kenapa.. ada apakah dengan Association Rules in Text Mining?

Can we use it to find semantic underlying the corresponding paragraph?

What about Inference Rules?

What are the problems in Text Understanding?