Many approaches have been proposed recently to identify the author of a given document. Thereby, one fact is often silently assumed: i.e., that the given document is indeed written by only author. For a realistic author identification system it is therefore crucial to at first determine whether a document is single- or multiauthored.

To this end, previous PAN editions aimed to analyze multi-authored documents. As it has been shown that it is a hard problem to reliably identify individual authors and their contribution within a single document (Author Diarization, 2016; Style Breach Detection, 2017), last year's task substantially relaxed the problem by asking only for binary decision (single- or multi-authored). Considering the promising results achieved by the submitted approaches (see the overview paper for details), we continue last year's task and additionally ask participants to predict the number of involved authors.

Given a document, participants thus should apply intrinsic style analyses to hierarchically answer the following questions:

Is the document written by one or more authors, i.e., do style changes exist or not?

If it is multi-authored, how many authors have collaborated?

All documents are provided in English and may contain zero up to arbitrarily many style changes, resulting from arbitrarily many authors.

The training set: contains 50% of the whole dataset and includes solutions. Use this set to feed/train your models.

Like last year, the whole data set is based on user posts from various sites of the StackExchange network, covering different topics and containing approximately 300 to 2000 tokens per document.

For each problem instance X, two files are provided:

problem-X.txt contains the actual text

problem-X.truth contains the ground truth, i.e., the correct solution in JSON format:

{ "authors": number_of_authors, "structure": [author_segment_1, ..., author_segment_3], "switches": [ character_pos_switch_segment_1, ..., character_pos_switch_segment_n, ] }

An example for a multi-author document could look as follows:

{ "authors": 4, "structure": ["A1", "A2", "A4", "A2", "A4", "A2", "A3", "A2", "A4"], "switches": [805, 1552, 2827, 3584, 4340, 5489, 7564, 8714] }

whereas a single-author document would have exactly the following form:

{ "authors": 1, "structure": ["A1"], "switches": [] }

Note that authors within the structure correspond only to the respective document, i.e., they are not the same over the whole dataset. For example, author A1 in document 1 is most likely not the same author as A1 in document 2 (it could be, but as there are hundreds of authors the chances are very small that this is the case). Further, please consider that the structure and the switches are provided only as additional resources for the development of your algorithms, i.e., they are not expected to be predicted.

To tackle the problem, you can develop novel approaches, extend existing algorithms from last year's task or adapt approaches from related problems such as intrinsic plagiarism detection or text segmentation. You are also free to additionally evaluate your approaches on last year's training/validation/test dataset (for the number of authors use the corresponding meta data).

This is the dataset for the Style Change Detection task of PAN 2022.

Task

The goal of the style change detection task is to identify text positions within a given multi-author document at which the author switches. Hence, a fundamental question is the following: If multiple authors have written a text together, can we find evidence for this fact; i.e., do we have a means to detect variations in the writing style? Answering this question belongs to the most difficult and most interesting challenges in author identification: Style change detection is the only means to detect plagiarism in a document if no comparison texts are given; likewise, style change detection can help to uncover gift authorships, to verify a claimed authorship, or to develop new technology for writing support.

Previous editions of the Style Change Detection task aim at e.g., detecting whether a document is single- or multi-authored (2018), the actual number of authors within a document (2019), whether there was a style change between two consecutive paragraphs (2020, 2021) and where the actual style changes were located (2021). Based on the progress made towards this goal in previous years, we again extend the set of challenges to likewise entice novices and experts:

Given a document, we ask participants to solve the following three tasks:

• [Task1] Style Change Basic: for a text written by two authors that contains a single style change only, find the position of this change (i.e., cut the text into the two authors’ texts on the paragraph-level),
• [Task2] Style Change Advanced: for a text written by two or more authors, find all positions of writing style change (i.e., assign all paragraphs of the text uniquely to some author out of the number of authors assumed for the multi-author document)
• [Task3] Style Change Real-World: for a text written by two or more authors, find all positions of writing style change, where style changes now not only occur between paragraphs, but at the sentence level.

All documents are provided in English and may contain an arbitrary number of style changes, resulting from at most five different authors.

Data

To develop and then test your algorithms, three datasets including ground truth information are provided (dataset1 for task 1, dataset2 for task 2, and dataset3 for task 3).

Each dataset is split into three parts:

1. training set: Contains 70% of the whole dataset and includes ground truth data. Use this set to develop and train your models.
2. validation set: Contains 15% of the whole dataset and includes ground truth data. Use this set to evaluate and optimize your models.
3. test set: Contains 15% of the whole dataset, no ground truth data is given. This set is used for evaluation (see later).

You are free to use additional external data for training your models. However, we ask you to make the additional data utilized freely available under a suitable license.

Input Format

The datasets are based on user posts from various sites of the StackExchange network, covering different topics. We refer to each input problem (i.e., the document for which to detect style changes) by an ID, which is subsequently also used to identify the submitted solution to this input problem. We provide one folder for train, validation, and test data for each dataset, respectively.

For each problem instance X (i.e., each input document), two files are provided:

1. problem-X.txt contains the actual text, where paragraphs are denoted by for tasks 1 and 2. For task 3, we provide one sentence per paragraph (again, split by ).
2. truth-problem-X.json contains the ground truth, i.e., the correct solution in JSON format. An example file is listed in the following (note that we list keys for the three tasks here):

   {
   "authors": NUMBER_OF_AUTHORS,
   "site": SOURCE_SITE,
   "changes": RESULT_ARRAY_TASK1 or RESULT_ARRAY_TASK3,
   "paragraph-authors": RESULT_ARRAY_TASK2
   }

   The result for task 1 (key "changes") is represented as an array, holding a binary for each pair of consecutive paragraphs within the document (0 if there was no style change, 1 if there was a style change). For task 2 (key "paragraph-authors"), the result is the order of authors contained in the document (e.g., [1, 2, 1] for a two-author document), where the first author is "1", the second author appearing in the document is referred to as "2", etc. Furthermore, we provide the total number of authors and the Stackoverflow site the texts were extracted from (i.e., topic). The result for task 3 (key "changes") is similarly structured as the results array for task 1. However, for task 3, the changes array holds a binary for each pair of consecutive sentences and they may be multiple style changes in the document.

   An example of a multi-author document with a style change between the third and fourth paragraph (or sentence for task 3) could be described as follows (we only list the relevant key/value pairs here):

   {
   "changes": [0,0,1,...],
   "paragraph-authors": [1,1,1,2,...]
   }

Output Format

To evaluate the solutions for the tasks, the results have to be stored in a single file for each of the input documents and each of the datasets. Please note that we require a solution file to be generated for each input problem for each dataset. The data structure during the evaluation phase will be similar to that in the training phase, with the exception that the ground truth files are missing.

For each given problem problem-X.txt, your software should output the missing solution file solution-problem-X.json, containing a JSON object holding the solution to the respective task. The solution for tasks 1 and 3 is an array containing a binary value for each pair of consecutive paragraphs (task 1) or sentences (task 3). For task 2, the solution is an array containing the order of authors contained in the document (as in the truth files).

An example solution file for tasks 1 and 3 is featured in the following (note again that for task 1, changes are captured on the paragraph level, whereas for task 3, changes are captured on the sentence level):

{
"changes": [0,0,1,0,0,...]
}

For task 2, the solution file looks as follows:

{
"paragraph-authors": [1,1,2,2,3,2,...]
}

This is the dataset for the shared task on Multi-Author Writing Style Analysis PAN@CLEF2025. Please consult the task's page for further details on the format, the dataset's creation, and links to baselines and utility code.

Task

The goal of the style change detection task is to identify text positions within a given multi-author document at which the author switches. Hence, a fundamental question is the following: If multiple authors together have written a text, can we find evidence for this fact; do we have a means to detect variations in the writing style? Answering this question belongs to the most difficult and most interesting challenges in author identification: Style change detection is the only means to detect plagiarism in a document if no comparison texts are given; likewise, style change detection can help to uncover gift authorships, to verify a claimed authorship, or to develop new technology for writing support.

Previous editions of the multi-author writing style analysis task aim at e.g., detecting whether a document is single- or multi-authored (2018), the actual number of authors within a document (2019), whether there was a style change between two consecutive paragraphs (2020, 2021, 2022), and where the actual style changes were located (2021, 2022). In 2022, style changes also had to be detected on the sentence level. The previously used datasets exhibited high topic diversity, which allowed the participants to leverage topic information as a style change signal. In the 2023 and 2024 editions of the writing style analysis task, special attention is paid to this issue.

We ask participants to solve the following intrinsic style change detection task: for a given text, find all positions of writing style change on the sentence-level (i.e., for each pair of consecutive sentences, assess whether there was a style change). The simultaneous change of authorship and topic will be carefully controlled and we will provide participants with datasets of three difficulty levels:

1. Easy: The sentences of a document cover a variety of topics, allowing approaches to make use of topic information to detect authorship changes.
2. Medium: The topical variety in a document is small (though still present) forcing the approaches to focus more on style to effectively solve the detection task.
3. Hard: All sentences in a document are on the same topic.

All documents are provided in English and may contain an arbitrary number of style changes. However, style changes may only occur between sentences (i.e., a single sentence is always authored by a single author and contains no style changes).

Data

To develop and then test your algorithms, three datasets including ground truth information are provided (easy for the easy task, medium for the medium task, and hard for the hard task).

Each dataset is split into three parts:

1. training set: Contains 70% of the whole dataset and includes ground truth data. Use this set to develop and train your models.
2. validation set: Contains 15% of the whole dataset and includes ground truth data. Use this set to evaluate and optimize your models.
3. test set: Contains 15% of the whole dataset, no ground truth data is given. This set is used for evaluation.

You are free to use additional external data for training your models. However, we ask you to make the additional data utilized freely available under a suitable license.