I need to develop an application to detect plagiarism imprint in source codes using a cognitive system, such as Watson. At this time I'm using Jplag (https://github.com/jplag/jplag) to analyze source codes and get tokenized files.
Here my problems, I have seen that Watson is a highly used system for unstructured files analysis. Instead, tokenized files are highly structured and can not be analyzed with most of the services provided by IBM Watson. For example, Watson Discovery can provide information about feelings or topics, within an email or publication and finds links between the files that make up the corpus. Unfortunately, Watson can not read the tokenized file for this purpose (or I do not know how to do it).
I have found Watson Analytics and it should provide correlation between numeric and structured files. But I can't reach the trial period so I can't test it.
I wonder if I can use Watson Analytics for my purpose. Should I insist in this way or did I miss some suggestion?
