Purpose
The purpose of this paper is to describe the development of an algorithm for realizing web crawlers that automatically collect dynamically generated webpages from the deep web.
Design/methodology/approach
This study proposes and develops an algorithm to collect web information as if the web crawler gathers static webpages by managing script commands as links. The proposed web crawler actually experiments with the algorithm by collecting deep webpages.
Findings
Among the findings of this study is that if the actual crawling process provides search results as script pages, the outcome only collects the first page. However, the proposed algorithm can collect deep webpages in this case.
Research limitations/implications
To use a script as a link, a human must first analyze the web document. This study uses the web browser object provided by Microsoft Visual Studio as a script launcher, so it cannot collect deep webpages if the web browser object cannot launch the script, or if the web document contains script errors.
Practical implications
The research results show deep webs are estimated to have 450 to 550 times more information than surface webpages, and it is difficult to collect web documents. However, this algorithm helps to enable deep web collection through script runs.
Originality/value
This study presents a new method to be utilized with script links instead of adopting previous keywords. The proposed algorithm is available as an ordinary URL. From the conducted experiment, analysis of scripts on individual websites is needed to employ them as links.
Survey of Techniques for Deep Web Source Selection and Surfacing the Hidden Web Content