Russian version

"... the act of plagiarism gives the impression that you wrote or thought something that you in fact borrowed from someone, and to do so is a violation of professional ethics.

Forms of plagiarism include the failure to give acknowledgement when repeating another's wording or particularly apt phrase, paraphrasing another's argument, and presenting another's line of thinking."

Joseph Gibaldi. MLA handbook for writers of research papers. Modern Language Association, 2013

Hack the Plagiarizer!

Hackathon for plagiarism detection in Russian texts

For the first time ever, a plagiarism detection for the Russian language hackathon is organized, co-located with the AINL conference.

Date: September 22-23, 2017

Location: Russia, Saint-Petersburg, ITMO, Lomonosova str. 9

Topic: Paraphrased plagiarism detection

Registration: or on the day of the event on-site

What is it about?

The Hackathon is focused on developing and evaluating algorithms for monolingual Russian plagiarism detection with the focus on scientific texts (academic plagiarism). The problem offered to the participants will be similar to the Text Alignment (TA) task evaluated at the PAN competitions; i.e., in a pair of texts, paraphrased or copy-pasted fragments taken from one text are to be found in a second text. For the task, the organizers provide training data. Participants are supposed to develop and train their approaches on this dataset.

The Hackathon opens with the talks by the Hackathon organizers. They will tell about the technologies used in plagiarism detection.

More details on the hackathon rules and workflow. (in Russian)

The Hackathon is free of charge for all participants of the main Conference. It means each participant should buy at least a 1-day ticket for the conference itself. Student discounts are available.


We encourage all the participants to register beforehand. Else, you can register on-site.


Measure\Team Baseline "Потом придумаем" "copycatboost" "HSE" "Смузи" "Nothing really mattress"
Plagdet Score 0.47 0.73 - - - -
Recall 0.39 0.75 - - - -
Precision 0.90 0.72 - - - -
Granularity 1.21 1.00 - - - -
F1 measure 0.54 0.73 - - - -