Analysis of Web Scraping Techniques to Get Keywords Suggestion and Allintitle Automatically from Google Search Engines

Abstract

Keyword research is one of the essential activities in Search Engine Optimization (SEO). One of the techniques in doing keyword research is to find out how many articles titles on a website indexed by the Google search engine contain a particular keyword or so-called "allintitle". Moreover, search engines are also able to provide keywords suggestion. Getting keywords suggestions and allintitle will not be effective, efficient, and economical if done manually for relatively extensive keyword research. It will take a long time to decide whether a keyword is needed to be optimized. Based on these problems, this study aimed to analyze the implementation of the web scraping technique to get relevant keyword suggestions from the Google search engine and the number of "allintitle" that are owned automatically. The data used as an experiment in this test consists of ten keywords, which each keyword would generate a maximum of ten keywords suggestion. Therefore, from ten keywords, it will produce at most 100 keywords suggestions and the number of allintitles. Based on the evaluation result, we got an accuracy of 100%. It indicated that the technique could be applied to get keywords suggestions and allintitle from Google search engines with outstanding accuracy values.

Analysis of Deep Learning Approach Based on Convolution Neural Network (CNN) for Classification of Web Page Title and Description Text

“…Precision, recall, and F1-score was also used to measure the detail performance of the proposed CNN architecture based on the distribution of the dataset. The precision was calculated by using an equation described in (5), the recall is calculated by using an equation described in (6), and the F1-score is calculated by using an equation described in (7). TP indicates true positive, FP indicates false positive, and FN indicates false negative.…”