Answer Evaluation of Short Descriptive Questions

Answer Evaluation of Short Descriptive Questions

DOI: 10.4018/978-1-7998-3772-5.ch005
OnDemand:
(Individual Chapters)
Available
$37.50
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

Reforms in the educational system emphasize more on continuous assessment. The descriptive examination test paper when compared to objective test paper acts as a better aid in continuous assessment for testing the progress of a student under various cognitive levels at different stages of learning. Unfortunately, assessment of descriptive answers is found to be tedious and time-consuming by instructors due to the increase in number of examinations in continuous assessment system. In this chapter, an attempt has been made to address the problem of automatic evaluation of descriptive answer using vector-based similarity matrix with order-based word-to-word syntactic similarity measure. Word order similarity measure remains as one of the best measures to find the similarity between sequential words in sentences and is increasing its popularity due to its simple interpretation and easy computation.
Chapter Preview
Top

Terminology Used

The terminology used in this chapter for finding similarity between answer content and solution key content is represented in Table 1 below.

Table 1.
Terminology used for question-answer evaluation
    Term                                  Meaning
SubjectPaper in different semesters of a course.
QTest paper of a subject with T questions shown as Q= {q1, q2..., qT}
AAnswer paper of a subject with T answers represented as A= {a1, a2..., aT}
SQuestion solution le of a subject with T solutions represented as S= {s1, s2..., sT}
aiAn answer ai consist of a set of j answer vectors represented as ai= {ai1, ai2..., aij}
siA solution si consist of a set of k solution vectors, represented as si= {si1, si2..., sik}
wi1, wi2..., wikPercentage of marks assigned to different solution vectors of solution, si= {si1, si2..., sik}
SIM (ai, si)A two-dimensional matrix for each question, with ai answer vectors and si solution vectors represented as ai×si with computed pair-wise similarity, say simaij,sik for every ai j answer vector and sik solution vector.
similarity (aij, sik)Pair-wise similarity, simai j,sik for every ai j answer vector & sik solution vector is represented as similarity(ai j,sik).
Theshold,δUser input threshold value to find the similarity
n(ai)Number of answer vectors in each ai
N(si)Number of solution vectors in each si
wc(ai j), wc(sik)wc(ai j) is the number of words in each answer vector, ai j and wc(sik) is the number of words in each solution vector, sik
cai j[w], csik[w]cai j[w] and csik[w] are the arrays of words common in answer vector ai j and solution vector sik.
vai j[u], vsik[u]vai j[u] and vsik[u] are the arrays of index numbers assigned to words in cai j[w] and csik[w] for u=1 to
wc(cai j[w])
instructor/
paper-setter
Carries out descriptive answer assessment.

Complete Chapter List

Search this Book:
Reset