International Workshop on Plagiarism Analysis,
Authorship Identification, and Near-Duplicate Detection (PAN)
held in conjunction with
The 30th Annual International ACM SIGIR Conference
23-27 July 2007, Amsterdam


Workshop Program

09:15-09:45Author Identification Using Imbalanced and Limited Training Texts
Efstathios Stamatatos [paper] [slides]
09:45-10:15Serial Sharers: Detecting Split Identities of Web Authors
Einat Amitay, Sivan Yogev, Elad Yom-Tov [paper] [slides]
10:15-11:00Coffee Break
11:00-11:30Forensic Authorship Attribution for Small Texts
Ol'ga Feiguina, Graeme Hirst [paper] [slides]
11:30-12:00Investigating Topic Influence in Authorship Attribution
George K. Mikros, Eleni K. Argiri [paper] [slides]
12:00-12:30Authors, Genre, and Linguistic Convention
Jussi Karlgren, Gunnar Eriksson [paper] [slides]
14:00-14:30Adaption of String Matching Algorithms for Identification of Near-Duplicate Music Documents
Matthias Robine, Pierre Hanna, Pascal Ferraro, Julien Allali [paper] [slides]
14:30-15:00Intrinsic Plagiarism Analysis with Meta Learning
Benno Stein, Sven Meyer zu Eissen [paper] [slides]

For the entire Proceedings see the publisher website.

Call for Papers

The workshop shall bring together experts and prospective researchers around the exciting and future-oriented topic of plagiarism analysis, authorship identification, and high similarity search. This topic receives increasing attention, which results, among others, from the fact that information about nearly any subject can be found on the World Wide Web. At first sight, plagiarism, authorship, and near-duplicates may pose very different challenges; however, they are closely related in several technical respects.

Plagiarism is the act of copying or including another author's ideas, language, or writing, without proper acknowledgment of the original source. Plagiarism analysis is a collective term for computer-based methods to identify a plagiarism offense. In connection with text documents we distinguish between corpus-based and intrinsic analysis: the former compares suspicious documents against a set of potential original documents, the latter identifies potentially plagiarized passages by analyzing the suspicious document with respect to changes in writing style. Such passages then can be used as a starting point for a Web search or for human inspection.

Plagiarism offense and plagiarism analysis can pertain to every kind of a digital media, e.g. text, program code, or images.

Authorship identification divides into so-called attribution and verification problems. In the authorship attribution problem, one is given examples of the writing of a number of authors and is asked to determine which of them authored given anonymous texts. If it can be assumed for each test document that one of the specified authors is indeed the actual author, the problem fits the standard paradigm of a text categorization problem. In the authorship verification problem, one is given examples of the writing of a single author and is asked to determine if given texts were or were not written by this author. As a categorization problem, verification is significantly more difficult than attribution. Authorship verification and intrinsic plagiarism analysis represent two sides of the same coin.

Near-duplicate detection is mainly a problem of the World Wide Web: duplicate Web pages increase the index storage space of search engines, slow down result serving, and decrease the retrieval precision. A naive solution to near-duplicate detection is the pairwise comparison of all documents. With specialized ''document models'', such as fingerprinting or locality sensitive hashing, a significantly reduced number of comparisons is necessary. These approaches provide efficient but incomplete solutions for solving the nearest neighbor problem in high dimensions.

Near-duplicate detection relates directly to plagiarism analysis: at the document level, near-duplicate detection and plagiarism analysis represent also two sides of the same coin. For a plagiarism analysis at the paragraph level, the same specialized document models (e.g. shingling, fingerprinting, hashing) can be applied, where a key problem is the selection of useful chunks from a document.

The development of new solutions for the outlined problems may benefit from the combination of existing technologies, and in this sense the workshop provides a platform that spans different views and approaches. The following list gives examples from the outlined field for which contributions are welcome (but not restricted to):

  • retrieval models for plagiarism analysis, authorship identification, and style analysis
  • software plagiarism, cross-language plagiarism, plagiarism in Web communities and social networks
  • NLP technologies for authorship identification and style analysis
  • knowledge-based methods for plagiarism analysis and authorship identification
  • handling proper citation

  • methods for identifying near-duplicate and versioned documents (for all kinds of contents, including text, source code, image, and music documents)
  • shingling, fingerprinting, and similarity hashing
  • hash-based search, high-dimensional search, approximate nearest neighbor search
  • efficiency issues and performance tradeoffs

  • tailored indexes for plagiarism analysis and near-duplicate detection
  • plagiarism analysis and near-duplicate detection on the Web
  • evaluation, building of test collections, experimental design and user studies

In particular, we encourage potential participants to present research prototypes and tools of their ideas.

The workshop addresses researchers, users, and practitioners from different fields: data mining and machine learning, document and knowledge management, semantic technologies, computer linguistics, social sciences, and information retrieval in general. We solicit contributions dealing with theoretical and practical questions of the development, use, and evaluation of theories and tools related to the workshop theme. Contributions will be peer-reviewed by at least two experts from the related field.

Submission Guideline

The workshop shall encourage the presentation of novel ideas in a possibly less formal way; however, we are still striving for high-quality contributions. Each contribution will be peer-reviewed by at least two reviewers, accepted papers will be included in the workshop proceedings. Depending on the quality of their contribution the author(s) of an accepted paper are invited to give a long or short presentation of their work. Moreover, we plan to invite the authors of the best papers to submit an overview article of their work for a special issue of an international journal, edited by the organizers with the aid of the program committee.

Submitted papers should be in the ACM Conference style, see the ACM template page, and may not exceed 6 pages. Submissions must generally be in electronic form using the Portable Document Format (PDF) or Postscript and mailed to The review is doubleblind; please anonymize your submission.

Important Dates

  • Jun 3, 2007:Postponed Deadline for paper submission
  • Jun 24, 2007:Notification to authors
  • Jul 3, 2007: Camera-ready copy due
  • Jul 27, 2007: PAN-07 Workshop

Organizing Committee

Program Committee

  • Shlomo Argamon, Illinois Institute of Technology
  • Yaniv Bernstein, Google Switzerland
  • Dennis Fetterly, Microsoft Research
  • Graeme Hirst, University of Toronto
  • Timothy Hoad, Microsoft
  • Heiko Holzheuer, Lycos Europe
  • Jussi Karlgren, Swedish Institute of Computer Science
  • Hans Kleine Büning, University of Paderborn
  • Moshe Koppel, Bar-Ilan University, Israel
  • Hermann Maurer, University of Technology Graz
  • Sven Meyer zu Eissen, Bauhaus University Weimar
  • Efstathios Stamatatos, University of the Aegean
  • Benno Stein, Bauhaus University Weimar
  • Özlem Uzuner, State University of New York
  • Debora Weber-Wulff, University of Applied Sciences Berlin
  • Justin Zobel, RMIT University