Semantic analysis system by using latent semantic analysis (LSA) algorithm on khutbah texts / Ammar Zuber

Friday Sermons is a pivotal part of the Friday prayers in the muslim’s community, as it purpose is to send out knowledge about the islam religion to its adherent. As nowadays, some muslims attend the Friday Prayers just to obey God instruction instead of willingly come to gain new knowledge. This is where the khutbah texts needs to have a certain level of quality and also new knowledge to ensure muslims does not go back home empty-handed. With the technology nowadays that comes with text analyser, JAIS should take advantage on this technology to use it on khutbah texts which is way more efficient than analysing it manually. In order to make it happen, this project aims to create a system to analyse khutbah texts semantically which analyse the documents meaning and its relationship to other documents compared to the standard text comparison. This project will be using Latent Semantic Analysis Algorithm to cope with comparisons of khutbah texts as it’s cover most of the requirement of the system. This project is developed by using Waterfall Model methodology as it’s the most direct model that is suitable for developing a system which consist of 5 phases which are Requirement Analysis, Design, Implementation, Testing and Maintenance. In the Requirement Analysis we managed to create the use case diagram and the system specification. Furthermore, in the design phase we constructed the system architecture, system flowchart, and the activity diagram. The implementation phase are done by developing the project based on the Python language and also the gensim package tool that are available for use. The testing phase is where two testing method were conducted which were the functionality testing and the benchmark testing. The key results that were found during the project is that it is possible to find the semantic features with the system can produce results based on the text meaning such as the system. With this information the system is able to produce results that can gives the user some useful information which can contribute in the later construction of khutbah texts. Although the system has some few limitation which it can hold back the user experience of the system, there are few recommendations which can be done in the future which can improve and enhance the system.