One’s writing originality in academic world becomes more and more questionable
along with the increasingly access to others’ writings due to files archiving technology
development today, especially over the internet. Therefore, a text similarity detection
system is required. Based on that problem, this research tries to provide the solution by
developing an application with the concept of text mining which implements cosine
similarity and Smith-Waterman algorithm to detect text similarity. Cosine similarity
serves to measure text similarity based on words occurrence, while Smith-Waterman
algorithm’s function is to calculate text similarity based on words sequence. Based on
this research test result, the developed application successfully detects text similarity
from very similar to very dissimilar pair of texts.