Natural Language Processing: An Alternative Approach to Analysis of Student's Summarization Strategies based on Heuristic Rules

2 Abstract: The rapid increase in the electronic information due global technology advancement as well as manual text usage has resulted in text information overload problem. This has leads to significant changes on text processing activities from the traditional way. Due to continous increase in both online and offline text, demand for text summarization has become the focus of huge interest and investment in both reserarch and commercial sectors. This research is aim to investigate various summarizing strategies in order to identified the commonly used strategies as well as problems their limitations in students' summary sentences. In this paper, we used heuristic rules to analyzed students' summaries. The results indicated that deletion is the commonly used summarization strategy among student's while topic sentence selection is the least. Also, it was found that grammatical error is the common problem affecting students' summary sentences followed by spelling mistakes. In addition, processes of sentences combination were identified and it was found out that apart from improper application of conjunction word and conjunction symbol, irrelevancy between combined sentences is the most frequent problem associated with sentences combination strategies. The results of this research can be used by teachers and institutions of learning to adapt their automatic summarization systems to improve teaching strategies which in turn enhanced student's summarization skills.