Online Automated Essay Grading System as a Web Based Learning (WBL) Tool in Engineering Education - Web-Based Engineering Education: Critical Design and Effective Tools - page 56

Intelligent Essay Assessor (Landauer, Foltz, &
Laham, 1998), Intellimetric (Elliot, 2001), a new
version of the Project Essay Grade (PEG, Page,
1994), and e-rater (Burstein et al., 1998).
Ellis Page set the stage for automated writing
evaluation (see the timeline in Figure 1). Rec-
ognizing the heavy demand placed on teachers
and large-scale testing programs in evaluating
student essays, Page developed an automated
essay-grading system called Project Essay Grader
(PEG). He started with a set of student essays that
teachers had already graded. He then experimented
with a variety of automatically extractable textual
features and applied multiple linear regressions
to determine an optimal combination of weighted
features that best predicted the teachers' grades.
His system could then score other essays using
the same set of weighted features. In the 1960s,
the kinds of features someone could automatically
extract from text were limited to surface features.
Some of the most predictive features Page found
included average word length, essay length in
words, number of commas, number of preposi-
tions, and number of uncommon words—the latter
being negatively correlated with essay scores.
In the early 1980s, the Writer's Workbench
tool (WWB) set took a first step toward this goal.
WWB was not an essay-scoring system. Instead,
it aimed to provide helpful feedback to writers
about spelling, diction, and readability. In addition

to its spelling program—one of the first spelling
checkers - WWB included a diction program
that automatically flagged commonly misused
and pretentious words, such as regardless and
utilize. It also included programs for computing
some standard readability measures based on
word, syllable, and sentence counts, so in the
process it flagged lengthy sentences as potentially
problematic. Although WWB programs barely
scratched the surface of text, they were a step in
the right direction for the automated analysis of
writing quality.
In February 1999, E-rater became fully op-
erational within ETS's Online Scoring Network
for scoring GMAT essays. For low-stakes writing-
evaluation applications, such as a Web-based
practice essay system, a single reading by an au-
tomated system is often acceptable and economi-
cally preferable. The new version of e-rater (V.2)
is different from other automated essay scoring
systems in several important respects. The main
innovations of e-rater V.2 are a small, intuitive,
and meaningful set of features used for scoring;
a single scoring model and standards can be used
across all prompts of an assessment; modeling
procedures that are transparent and flexible, and
can be based entirely on expert judgment.
Figure 2 shows a popular common frame
work of the automated essay grading systems.
Most of the modern systems train the system

Figure 1. A timeline of research developments in writing evaluation