Language models and fusion for authorship attribution

Detection of changes in literary writing style using N-grams as style markers and supervised machine learning

The analysis of an author’s writing style implies the characterization and identification of the style in terms of a set of features commonly called linguistic features. The analysis can be extrinsic, where the style of an author can be compared with other authors, or intrinsic, where the style of an author is identified through different stages of his life. Intrinsic analysis has been used, for example, to detect mental illness and the effects of aging. A key element of the analysis is the style markers used to model the author’s writing patterns. The style markers should handle diachronic changes and be thematic independent. One of the most commonly used style marker in extrinsic style analysis is n-gram. In this paper, we present the evaluation of traditional n-grams (words and characters) and dependency tree syntactic n-grams to solve the task of detecting changes in writing style over time. Our corpus consisted of novels by eleven English-speaking authors. The novels of each author were organized chronologically from the oldest to the most recent work according to the date of publication. Subsequently, two stages were defined: initial and final. In each stage three novels were assigned, novels of the initial stage corresponded to the oldest and those at the final stage to the most recent novels. To analyze changes in the writing style, novels were characterized by using four types of n-grams: characters, words, Part-Of-Speech (POS) tags and syntactic relations n-grams. Experiments were performed with a Logistic Regression classifier. Dimension reduction techniques such as Principal Component Analysis (PCA) and Latent Semantic Analysis (LSA) algorithms were evaluated. The results obtained with the different n-grams indicated that all authors presented significant changes in writing style over time. In addition, representations using n-grams of syntactic relations have achieved competitive results among different authors.

Authorship Attribution in Bangla Literature (AABL) via Transfer Learning using ULMFiT

Authorship Attribution is the task of creating an appropriate characterization of text that captures the authors’ writing style to identify the original author of a given piece of text. With increased anonymity on the internet, this task has become increasingly crucial in various security and plagiarism detection fields. Despite significant advancements in other languages such as English, Spanish, and Chinese, Bangla lacks comprehensive research in this field due to its complex linguistic feature and sentence structure. Moreover, existing systems are not scalable with the increasing number of authors, and performance drops with the small number of samples per author. In this paper, we propose the use of Average-Stochastic Gradient Descent Weight-Dropped Long Short-Term Memory (AWD-LSTM) architecture and an effective transfer learning approach that addresses the problem of complex linguistic features extraction and scalability for authorship attribution in Bangla Literature (AABL). We analyze the effect of different tokenization, such as word, sub-word, and character level tokenization, and demonstrate the effectiveness of these tokenizations in the proposed model. Moreover, we introduce the publicly available Bangla Authorship Attribution Dataset of 16 authors (BAAD16) containing 17,966 sample texts and 13.4+ million words to solve the standard dataset scarcity problem and release six variations of pre-trained language models for use in any Bangla NLP downstream task. For evaluation, we used our developed BAAD16 dataset as well as other publicly available datasets. Empirically, our proposed model outperformed state-of-the-art models and achieved 99.8% accuracy in the BAAD16 dataset. Furthermore, we showed that the proposed system scales much better with the increasing number of authors, and performance remains steady even with few training samples.

UrduAI: Writeprints for Urdu Authorship Identification

The authorship identification task aims at identifying the original author of an anonymous text sample from a set of candidate authors. It has several application domains such as digital text forensics and information retrieval. These application domains are not limited to a specific language. However, most of the authorship identification studies are focused on English and limited attention has been paid to Urdu. However, existing Urdu authorship identification solutions drop accuracy as the number of training samples per candidate author reduces and when the number of candidate authors increases. Consequently, these solutions are inapplicable to real-world cases. Moreover, due to the unavailability of reliable POS taggers or sentence segmenters, all existing authorship identification studies on Urdu text are limited to the word n-grams features only. To overcome these limitations, we formulate a stylometric feature space, which is not limited to the word n-grams feature only. Based on this feature space, we use an authorship identification solution that transforms each text sample into a point set, retrieves candidate text samples, and relies on the nearest neighbors classifier to predict the original author of the anonymous text sample. To evaluate our solution, we create a significantly larger corpus than existing studies and conduct several experimental studies that show that our solution can overcome the limitations of existing studies and report an accuracy level of 94.03%, which is higher than all previous authorship identification works.

Cross-Domain Authorship Attribution Using Pre-trained Language Models

Authorship attribution attempts to identify the authors behind texts and has important applications mainly in cyber-security, digital humanities and social media analytics. An especially challenging but very realistic scenario is cross-domain attribution where texts of known authorship (training set) differ from texts of disputed authorship (test set) in topic or genre. In this paper, we modify a successful authorship verification approach based on a multi-headed neural network language model and combine it with pre-trained language models. Based on experiments on a controlled corpus covering several text genres where topic and genre is specifically controlled, we demonstrate that the proposed approach achieves very promising results. We also demonstrate the crucial effect of the normalization corpus in cross-domain attribution.