Effects of Using Arabic Web Pages in Building Rank Estimation Algorithm for Google Search Engine Results Page

Abstract

Search Engine Optimization (SEO) aims to improve a website's reputation and user experience. Without effective SEO strategies, it requires significant investment in paid advertisements. Search Engines (SEs) use algorithms to rank results, assessing on-page and off-page factors for relevance. Machine learning techniques have been used to build classifiers for estimating page rank. However, no research has compared rank estimation with other languages or analyzed the effects of different languages on performance or differences between SEO factors. The study aims to improve rank estimation algorithms for Arabic web pages on desktop devices using a new multi-category dataset from Google Search Engine Results Page (SERP). The experimental findings suggest that Arabic web pages are more suitable than English ones for training a model to estimate the ranking of Arabic web pages. Machine learning models were applied to two datasets. SE scraping was used to collect URLs, descriptions, and other data from the Google SE. Data preprocessing steps were taken before using the datasets for rank estimation algorithms. Experiments were conducted to assess the implications of using Arabic and English web page datasets