English Reading Assistant software

[cools...@hotmail.com]

The idea of an "English reading assistant" software program came into
my mind when I was thinking of a solution to what BabelCode is weak at
- helping non-English speakers comprehend natural English text.

We already know that a BabelCode author can create natural English
documents without any knowledge of English grammar or vocabulary. But
the reversed thing isn't true; he still has problems reading a natural
English document.

My solution to this problem is not letting the computer give several
likely interpretations for the reader to choose. That would be way too
laborious, even worse than letting the reader go to learn English. But
learning English alone won't solve the problem very soon. Even if the
user have learned basic English grammar and vocabulary, there are still
too many English idioms and new words which stop the reader from
understanding an English document well. So I get an idea of making an
English reading assistant software program. The software mainly does
two things:

(1) Translate the meaning of those not-so-common words and suggest the
most likely word senses for those ambiguous words.

(2) Discover idiomatic usages and tell the user what they really mean.

We can use knowledge-based and statistical methods which are commonly
used in NLP systems to implement the above 2 tasks.

The program can display its suggestions (tips for the reader) on a
"sidebar" on the right of the main English document. The occurrences it
is making tips for are underlined.

BTW, Microsoft Research also made an "English writing wizard" (EWW)
program released within its Office suite. Its idea is a complement to
my idea.

Yao Ziyuan
in...@babelcode.org

[yaoz...@gmail.com]

In extension to this original idea:

(1) "English" can be extended to all possible foreign languages;
(2) The program should outline essential word usages as used in the
main document. Usage information is presented as described in the
Usenet message "LingoX: Foreign Language Writing Aid Software", e.g.
"v. search [n. someplace] [for n. someobject]".
(3) The primary text retrieval method may be "copying to the
Clipboard". A "Checkpoints" window will display the usage checkpoints
that are found by the program from the retrieved text. For major target
applications such as Internet Explorer, specialized automatic text
retrieval modules may be developed to achieve a better reading
experience.

Yao Ziyuan
cools...@hotmail.com

[yaoz...@gmail.com]

The program can keep track of which checkpoints are already known to
the user, in order to always show new checkpoints. The user should also
be able to set filters to receive only a certain type of checkpoint.
Yao Ziyuan
cools...@hotmail.com

[yaoz...@gmail.com]

[yaoz...@gmail.com]

The reading assistant program should also capture text documents in the
user's native language and extract checkpoints from them, because the
user is even more interested in documents written in native language,
therefore enabling better incentive.

Yao Ziyuan

[yaoz...@gmail.com]

Learning phrases (e.g. subject-predicate phrases, verb-complement
phrases, modifier-headword phrases, noun phrases, adverbial phrases) is
important to a student, but this learning activity can't be elegantly
assisted in a writing process (there are three ways of doing this: 1)
the user has to look up for phrases inputted in his native language,
which requires motivation; 2) the computer waits until the user inputs
a "Chinglish" phrase in his writing and then suggests him the natural
usage, which forces the user to make mistakes before he could learn the
correct usages; 3) the computer waits until it has collected enough
information to make a least intrusive prompt AND the user has not made
a full mistake, which is impractical; Microsoft's EWW is exactly doing
this way, resulting prompts too intrusive.).

So I think such learning activity (phrase translation knowledge
acquisition) should be best aided in the READING process (both for
NATIVE LANGUAGE READING and FOREIGN LANGUAGE READING). When the user is
reading some information interesting to him, the computer displays (on
a right-hand-side "Sidebar" on the Desktop) bilingual pairs of
essential phrases (and new words and idiomatic sentences) extracted
from the document the user is currently reading.