Paper Contents
Abstract
Google searches play a crucial role in providing accurate and relevant information, particularly in the domain of politics. Enhancing Google searches for political matters is essential due to the complex nature of politics and the need for reliable and up-to-date information. This paper proposes an approach to improve information retrieval in Arabic language sources by addressing the challenges associated with the language's complexity. The proposed approach includes several steps. First, tokenization divides the user's query into individual word segments, allowing for further processing based on the selected domain. Unification ensures consistent representation of Arabic letters by addressing the variations caused by diacritics. Stop-words and special characters that offer little semantic value are removed to improve precision. The approach incorporates light stemming using the "Khoja" stemmer, which generates relevant terms without over-generating them. Term generation expands the range of terms by using a mechanism proposed by "Sarf" and selecting the top ten terms with the highest term frequency on Google. Finally, the query is updated on the backend to increase the bag of words for evaluation. To evaluate the effectiveness of the proposed approach, precision is used as the primary metric. Google Search Engine (SE) serves as the benchmark for comparison, considering its efficiency in Arabic language information retrieval. The precision values of applied queries related to the Politician domain are recorded per page, both in their original plain form and after being updated using the proposed approach. The results demonstrate that the proposed approach improves precision compared to the original plain queries. For instance, precision increases from 0.40 to 0.65 for Q1 and from 0.65 to 0.95 for Q3. These findings highlight the effectiveness of the approach in enhancing the retrieval of relevant documents in Arabic language information retrieval systems. The systematic process presented in this research contributes valuable insights for improving the performance of information retrieval systems in the Arabic language. Further research can focus on refining and optimizing the approach, exploring its applicability to other domains, and addressing any remaining challenges to ensure its effectiveness in real-world scenarios.
Copyright
Copyright © 2023 Mansour Ali Ahmad Al-helalat. This is an open access article distributed under the Creative Commons Attribution License.
