Mohd Zamri Murah

A malay stemmer for jawi characters

The Malay language may be written using either Roman or Jawi characters. Most Malay stemmers cover only Roman (Rumi ) affixes. This paper proposes a stemmer for Jawi characters using two sets of rules in Jawi: one set of rules is used to stem various forms of derived words, and another set is used to replace the use of a dictionary by producing the root word for each derivative. This stemmer has been tested using 1185 derived words consisting of prefix, circumfix, suffix, and infix. The results show that 84.89% of Jawi root words have been successfully stemmed.

Spelling error detector rule for Jawi stemmer

Stemmer is important especially for information and document retrieval. It can also help to reduce the size of the dictionary. Normally Malay stemmers need to have a root word dictionary to increase the stemmers accuracy. In Jawi stemmer, we use Jawi spelling error rule to detect whether the program produces the correct stemmed word after all possible affixes have been removed. Jawi spelling error rule has been tested using 3018 data in Jawi with two syllables root word and the result was compared manually. The result shows 97.8% accuracy of Jawi spelling word with two syllables which have been checked correctly using the ‘spelling error detector rule’.

Part-of-Speech for Old Malay Manuscript Corpus: A Review

Research in Malay Part-of-Speech (POS) has increased considerably in the past few years. From the literature, POS are known as the first stage in automated text analysis and the development of language technologies can scarcely begun without this initial phase. Malay language can be written in Roman or Jawi. Three different spelling between Roman and Jawi make this study essential. In this paper, we highlighted the problem and issues related to Malay language, POS general framework, POS approaches and techniques. POS at basis was introduced to get information from Old Malay Manuscripts that contain important information in various spheres of knowledge. Promising result for the auto-tagging of Malay written in Jawi is expected.

Evaluation of Microsoft speech recognition in controlling robot soccer

This paper presents the evaluation of speech recognition. The main objective of the design is to develop new algorithm for control the motion of the robot such as forward, reverse, left, right and stop. The proposed method is divided to speech recognition and analytical engine. This system has been tested in two situation i.e. silent and noisy condition. It has been proved this system work very effectively in silent condition.

License plate localization based on Kapur optimal multilevel threshold

A license plate localization system is useful for many applications. Due to ambient of lighting in three distinct situation which are morning, afternoon and night causing difficulty to search optimum threshold value in each situation. This research uses global thresholding approach by using Kapur entropy multilevel threshold based on Patch-Levy Bees Algorithm (PLBA). As a result, the system properly localize and identify number plate in the image by using proposed segmentation image. From the experiment, proposed method are achieve accuracy rates to 67.68%, 90.71%, 24.34% respectively for morning, afternoon and night dataset.

Tokenizer for the Malay language using pattern matching

Tokenization is a fundamental task focused on text processing. Among other tasks, the segmentation process is used to identify information units, such as sentences and words. In this paper, we discuss the Natural Language ToolKit (NLTK) tokenizer as a step to manipulate patterns within text. The purpose of this work is to build up Natural Language Processing (NLP) base for Jawi corpus. A series of experiments was performed, to validate the corpus and fulfill the requirement of the Jawi script tokenizer, with the promising results. Based on these promising results, the token will be used for tagging process.

The Effectiveness of a Jawi Stemmer for Retrieving Relevant Malay Documents in Jawi Characters

The Malay language has two types of writing script, known as Rumi and Jawi. Most previous stemmer results have reported on Malay Rumi characters and only a few have tested Jawi characters. In this article, a new Jawi stemmer has been proposed and tested for document retrieval. A total of 36 queries and datasets from the transliterated Jawi Quran were used. The experiment shows that the mean average precision for a “stemmed Jawi” document is 8.43%. At the same time, the mean average precision for a “nonstemmed Jawi” document is 5.14%. The result from a paired sample t-test showed that the use of a “stemmed Jawi” document increased the precision in document retrieval. Further experiments were performed to examine the precision of the relevant documents that were retrieved at various cutoff points for all 36 queries. The results for the “stemmed Jawi” document showed a significantly different start, at a cutoff of 40, compared with the “nonstemmed Jawi” documents. This result shows the usefulness of a Jawi stemmer for retrieving relevant documents in the Jawi script.

Implementation of Buckwalter transliteration to Malay corpora

Assigning lexical categories to words is an important step in the automated analysis of a text. Modern Natural Language Processing (NLP) algorithms are based on machine learning; learn rules automatically through the analysis of large corpora of typical real world examples. The Buckwalter transliteration has become a standard to be followed in natural language processing research community that works on Arabic. In this paper, we discuss the encoding in Malay language corpus written in Jawi. The purpose of this work is to conform and standardize the corpora between the similar characters. Four different letters with the Arabic language identified and new defined Buckwalter symbols were assigned to the letters. Collections of 114 chapter in al-Quran translated in Jawi has been used as a corpora. The similar corpora between Jawi and Arabic language will be manipulated to determined out-of-vocabulary problem (OOV) in POS-tags.