Plagiarism check - music teacher question
Hi guys, I'm a high school music teacher and I have had students hand in compositions that I suspect may be plagiarised. I can't search by title because obviously they would have changed the titles of their pieces. Is there some way I can do a plagiarism check on here? I've been trying to search through the pieces on here but I haven't had any luck yet!
I've uploaded the piece in question - date of creation is 30th May 2016.
| Attachment | Size |
|---|---|
| Cool_Beans - name removed.mscz | 29.71 KB |
Comments
Ask Led Zeppelin's lawyer. Or, have a look at websites relating to music identification algorithms. Unfortunately, whilst there are algorithms to compare pieces of music you will still have to input them or visually search scores yourself. Unless anyone had developed a Google search of music, of course.
A web search for "song identification" yields quite a few results. Hopefully you can find something that works for you.
AFAIK, there is no incipit search function available for the MuseScore server as there is for RISM. So if you suspect one of your students has stolen someone else's work from here, you're going to have to look for it by eye, a somewhat daunting task given the sheer number of new compositions posted here every day.
OTOH, a lot of the compositions posted publicly here aren't much better than high-school level anyway, so there'd be little advantage in stealing them except for a student who procrastinated till the last second and needed something, ANYthing to turn in. (Professional composers tend to mark their scores private, which means no one but the composer--and people to whom he gives the appropriate log-in information--can see or download them.)
What you can do if your students are submitting their scores in MuseScore format (mscz), is click on File>Info and look at the source and date created. If someone downloaded a score they found on the server and were not 'wise' enough to modify the information embedded in the underlying file, you'll see it right there. The date created will give clues as to whether the work was done recently or many months ago, and the source line in the info panel provides a link to the score as stored on the server. If you're in doubt about the bona fides of the student's work, go to that page and see who the user is that created it.
In reply to AFAIK, there is no incipit by Recorder485
Brilliant! Though I recall you're sticking with last year's MuseScore—now it's File > Score Properties... @mhann, if there's anything in the "Source" field, copy and paste that into your browser's address bar to see where it came from.
In reply to Brilliant! Though I recall by Isaac Weiss
Ahhhh! I saw a passing reference to 'Score Properties' in a post the other day and it puzzled me because I couldn't find that menu item anywhere. Thanks for educating me.
I suppose you could post the suspicious score(s) and there is a chance that someone will recognize it and say "Hey, I wrote that!".
In reply to I suppose you could post the by underquark
I considered that. I'll start with one of them but I'll remove the name of the student:
In reply to I considered that. I'll start by mhann47
That piece was created on May 30th, 2016, on an Apple MacIntosh computer, using MuseScore 2.0.1, which is two revisions older than the present version and dates from over a year ago, IIRC. How long has this student been using MuseScore? Can you compare this information (Apple, 2.0.1) with the properties embedded in other submissions by the same student?
There is no source shown in the Score Properties/Info panel, so either the score was never saved on the MuseScore server, or the student wiped that information off the file...which begs the question of why he would do that if it were really his work.
No comment on the piece itself except to say it's the sort of thing a good high-school guitarist might play by ear, but as to notating what he can play by ear, that is another story. That is fairly complex notation work, both rythmically and harmonically.
In reply to That piece was created on May by Recorder485
.
You could ask the students to talk you through the piece in terms of why they chose to do things the way they did. If any of them appear unfamiliar with the piece or the techniques used then that could be an indication that it was plagiarised.
In reply to You could ask the students to by shoogle
While I'm not able to do it, I think this might be a candidate for a plugin, which would take a piece submitted and compare it to uploaded scores.
In reply to While I'm not able to do it, by xavierjazz
There are currently over a hundred and eighty-seven thousand publicly viewable scores stored on the MuseScore server (as of ten minutes ago: 18,710 pages containing ten scores each, plus four more). No doubt a comparison algorithm could be worked out, but running such a program on a data-base that large would take a massive amount of computing power and a very fast internet connection, unless you didn't care how long it took to run. Think of how long it would take just to download 187,000 scores of 100kb (slightly over a month, at 15 seconds per score), and then add in the time needed to parse them all.
You'd need the computing power of a major search engine operator, like Google, and I'm pretty sure the app would have to run server-side.
A much more reasonable project would be to create an incipit search function similar to what RISM has on their site, and add it to MuseScore.com.
In reply to There are currently over a by Recorder485
Thanks.
In reply to Thanks. by xavierjazz
Something like
http://www.midomi.com/
http://www.watzatsong.com/
http://www.folktunefinder.com/
and the like could perhaps help? (but research would be difficult for the large number of scores)
In reply to Something by Shoichi
Midomi looks moderately interesting, but requires a person to be able to sight-sing accurately (or, I suppose, to play the theme on an instrument). Watzatsong requires creating a user account, so I didn't go further. Folk tune finder isn't functioning in my browser; I can't input any notes.
Here is a link to the RISM incipit search tool. https://opac.rism.info/metaopac/start.do?View=rism&SearchType=2&Languag…
In reply to Midomi looks moderately by Recorder485
It was just an idea. If something like that was limited to act on MuseScore.com database (+/- 187 400 scores) it could serve to find a 'lost' track.
In reply to It was just an idea. If by Shoichi
The RISM tool works by searching for the theme entered in the INCIPITS for each score in their database. It does not take an upload of an entire score and attempt to compare it to whole scores in its database. So the size of the searched material is MUCH smaller than the total MuseScore database. When music librarians submit editions (whether autograph or copyist's MSS, or engraved publications) to RISM for listing, an incipit is manually created and linked to the page. That consists of the first few measures (for each movement, when a multi-movement work). That's all that's usually necessary to locate a piece by theme.
For the purposes of searching the MuseScore database for plagerised work, an incipit search would probably be more than sufficient. It would be somewhat easier, I suspect, to create such a tool because all the scores on MuseScore.com are already in a digital format that can be parsed accurately by an app. RISM has to create those incipits manually; MuseScore would simply have to define in the code how many measures to search.
In reply to The RISM tool works by by Recorder485
For refence: https://musescore.org/en/node/116261#new