Information Technology Reference
In-Depth Information
Figure 3. Proposed framework of the AEG system with local language engines
After submission they get the final projected score.
Hence these engines help the students to learn
better English. To make these engines effective
the system I strained with a good number of lo-
cal words, which are very much used in normal
English (spoken English, news paper English).
To make a proper collection of local words the
local English news papers are used as a source.
As for example - to make the engine working in
Andhrapa Pradesh it is trained on collection of
local words used in the news papers like Deccan
Chronicle, Hindu (AP edition), Times of India (AP
edition) etc., collected over last couple of years.
It is found that this specific region's English is
influenced by Telugu and Hyderabadi Hindi (a
good mixing of Hindi and urdu).
scores and poor grammar, usage and mechanics
scores compare to other students. More over lo-
cal languages influence them. Serious work in
the area of AEG can bring significant changes in
this direction and also can give a new shape to
Indian NLP & Machine Learning research work.
Future plans - In near future the following
things will be taken into consideration so that
some solutions can be given as - Solution for ma-
chine translated essays (how to recognize them?),
Capturing the mental status of the student writing
essay (psychometric models will be considered).
Detection of Anomalous Essays
rEfErEncES
Bloom, B. S. (1956).
Taxonomy of educational
objectives: The classification of educational
goals. Handbook I, Cognitive domain
. New York:
Longmans, Green.
6. concLuSIon
In his paper 'Region Effects in AEG & human
discrepancies of TOEFL score' Attali (2005) men-
tioned Asian Students show higher organization
Search WWH ::
Custom Search