40 research outputs found
NL scoring technique for the assessment of learners' understanding
No full textAssessment is an important component of learning and it also noted that many academic examination make heavy used of short answers. This assessment can be a tedious task. However, there are not many computer-based assessment tools due to limitations in computerized marking technology. Our
research attempts to address this limitation by introducing a technique to evaluate short free text answer. It is based on a hybrid approach that combines natural language processing, node-link representation and information theory. A textual answer is converted into a node link representation to extract the hidden knowledge structure. We then apply excess entropy to compute the amount of known
information for each model and later compute the score accordingly. Results show that the proposed technique can be applied for the assessment of learners’ understanding in tertiary and school level for science domain
Part-of-Speech in a Node-Link Scoring Techniques for Assessing Learners’ Understanding
No full textAbstractAssessment is the process of which the quality of the candidate achievements can be judged. Automated assessment tools can be useful assistants to human examiners. This paper discusses a work on an automated assessment tool that applied a node-link analysis technique. The tool is able to assess short sentences answers. It uses Part-of-speech as important criteria during the node analysis process. Results have shown that using more relevant Part-of-speech gives a more reliable assessment results in judging a learners’ understanding on a particular subject matter
Modelling learners’ understanding through Node-link analysis and scoring
No full textAssessment is an important component of learning and it also noted that many examinations make heavily used of short answers. This assessment can be a tedious task. This research work proposed a technique that models the teacher's and learner's knowledge to evaluate short free text answer. It is based on a hybrid approach that combines natural
language processing, a node-link as a representation technique and the application of information theory in
the measurement of learners' understanding. Pearson correlation and exact-and-adjacent agreement have been used as a measurement instrument. Results show that the proposed technique can be applied for the assessment of learners' understanding in tertiary and school level for computer science and science domain
Deep Learning Approach for cognitive competency assessment in Computer Programming subject
Get PDFThis research examines the competencies that are essential for an lecturer or instructor to evaluate the student based on automated assessments. The competencies are the skills, knowledge, abilities and behavior that are required to perform the task given, whether in a learning or a working environment. The significance of this research is that it will assist students who are having difficulty learning a Computer Programming Language course to identify their flaws using a Deep Learning Approach. As a result, higher education institutions have a problem with assessing students based on their competency level because; they still use manual assessment to mark the assessment. In order to measure intelligence, it is necessary to identify the cluster of abilities or skills of the type in which intelligence expresses itself. This grouping of skills and abilities referred to as "competency". Then, an automated assessment is a problem-solving activity in which the student and the computer interact with no other human intervention. This review focuses on collecting different techniques that have been used. In addition, the review finding shows the main gap that exists within the context of the studied areas, which contributes to our key research topic of interest
Assessment of learners’ understanding: an experimental result
No full textAssessing leaner’s answers is very time consuming for educators and limits them to be involved in other activities. In many cases, exam papers comprise of questions that require learners to write at least one or two sentences to express their understanding. However, there are not many computer-based assessment tools due to limitations in computerized marking technology. Our research attempts to address this limitation by introducing a technique to evaluate short free text answer. It is based on a hybrid approach that combines natural language processing, information extraction and artificial intelligence. A textual answer is converted into a node link representation to extract the hidden knowledge structure. We then apply excess entropy to compute the amount of known information for each model and later compute the score accordingly
Competency assessment of short free text answers
Get PDFThe increased adoption of competency-based
education has posed the need of an automated competency
assessment. Most of the existing assisted assessment does not
cater for competency assessment. The high percentage on the use
of short free text answer as competency assessment shows that
the need of the competency assisted assessment is urgent. This
paper studies on the need and also review on existing assisted
assessments focusing on short free text answer. A Node Link
(NL) Scoring technique is proposed as an alternative automated
solution to assess learners’ competency in short free text answers.
Keyword: competency assessment; short free text answers; Node
Link Scoring techniqu
Assessing learner’s understanding via node-link analysis: an information theory approach
No full textCognition or thinking involves mental activities or thought process. The complexness of thought process is represented by the complexness of textual information that consists of more distinct pieces of information. Textual information which is in the form of language and symbol are means in representing and communicating information, experiences and
ideas. In our research, the textual information is converted into a node-link representation and the development of the node and link is based on the word co-occurrence and meaning. This process enables us to uncover the hidden knowledge structure. An expert model is used as a referent to enable a precise assessment on learner’s understanding or known information on a specific subject. Initial results show that the more complex the relationships among the terms (pieces of information), the higher the amount of bits
generated, as a measurement of the learner’s understanding
