In general, hadith consists of isnad and matan (content). Matan can be separated into several components for example a story, main content, and some additional information. Other texts besides main content, such as isnad and story can interfere the retrieval process of relevant documents because most users typically use simple queries. Thus, in this paper, we proposed a Named Entity Recognition (NER) component weighting model in improving the Indonesian hadith retrieval system. We did 3 test scenarios, the first scenario (S1) did not separate the hadith into several components, the second scenario (S2) separated the hadith into 2 components, isnad and matan, and the third scenario separated the hadith into 4 components, isnad, background story, content, and additional information. From the experimental results, it is found that the TF-IDF with rocchio algorithm in query expansion outperforms DocVec. Also, separation and weighting of the hadith components affect the retrieval performance because isnad can be considered as noise in a query. Separation of 2 separate components had the best overall results in general although 4 separate components showed better results in some cases with precision up to 100% and 70% recall.

No abstract

Computerized adaptive testing (CAT) is a context-based adaptive assessment. However, the assessment result may not be valid because the examinee might cheat or guess the answers. Although there are many guessing detection methods, there are not many discussions about their implementation into CAT. Therefore, this paper presents an example of a modification of an existing software so the newly modified software can detect guessed answers and be able to select questions adaptively.The system can detect assuming behavior by recording the examinee's answer time.Also, the designed system can like questions adaptively by connecting Fuzzy logic, which calculates what level the question should select for the next iteration. The system is responded well by elementary and college students. A total of 56.6% felt the system was straightforward to use. The detection methods can detect guessing behavior of about 73%. However, the system's sensitivity is low if the method is forced to classify answers which answered in a long response time / general guessing. Nevertheless, when we limit the data classified within 10s response time (rapid-guessing), the method's sensitivity rises to 68.78%.

Assessment is one benchmark in measuring students’ abilities. However, assessment results cannot necessarily be trusted, because students sometimes cheat or even guess in answering the questions. Therefore, to obtain valid results, it is necessary to separate valid and invalid answers by considering rapid-guessing behaviour. We conducted a test to record exam log data from undergraduate and postgraduate students to model rapid-guessing behaviour by determining the threshold response time. Rapid-guessing behaviour detection is inspired by the common k-second method. However, the method flattens the application of the threshold, thus allowing misclassification. The modified method considers item difficulty in determining the threshold. The evaluation results show that the system can identify students’ rapid-guessing behaviour with a success rate of 71%, which is superior to the previous method. We also analysed various aggregation techniques of response time and compared them to see the effect of selecting the aggregation technique.

No abstract