Improving Plagiarism Checker Throughput with Apache Storm

At present, accessing and publishing new knowledge are much easier than they were in the past. This has given rise to a “copy-and-paste” culture, which is a violation of intellectual property. A plagiarism checker can help to identify such incident. Unfortunately, existing plagiarism checkers still do not offer services that handle many users at once like what Google’s do. We conjecture that there is a bottleneck that they still need to overcome before making their adoption more widespread. This paper proposes a framework to improve the throughput of existing plagiarism checkers to tackle the bottleneck in question. Our framework leverages stream processing via Apache Storm. It can be tailored to meet throughput specification with additional computational resources that grow linearly with the increase in throughput.