A graph-based multi-level linguistic representation for document understanding

Graph-Based Siamese Network for Authorship Verification

In this work, we propose a novel approach to solve the authorship identification task on a cross-topic and open-set scenario. Authorship verification is the task of determining whether or not two texts were written by the same author. We model the documents in a graph representation and then a graph neural network extracts relevant features from these graph representations. We present three strategies to represent the texts as graphs based on the co-occurrence of the POS labels of words. We propose a Siamese Network architecture composed of graph convolutional networks along with pooling and classification layers. We present different variants of the architecture and discuss the performance of each one. To evaluate our approach we used a collection of fanfiction texts provided by the PAN@CLEF 2021 shared task in two settings: a “small” corpus and a “large” corpus. Our graph-based approach achieved average scores (AUC ROC, F1, Brier score, F0.5u, and C@1) between 90% and 92.83% when training on the “small” and “large” corpus, respectively. Our model obtain results comparable to those of the state of the art in this task and greater than traditional baselines.

Using weighted directed graphs for identification of flow of emotions in poems

Poem is a spontaneous flow of emotions. There are several emotion detection systems to identify emotions from speech, gestures, and text (blogs, newspapers, stories and medical reports). Since such systems do not exist for poetry, we take the first step in building a system to recognize emotions in poetry by constructing a benchmark corpus, the PERC (Poem Emotion Recognition Corpus), of poems written by Indian poets in English. In this research a novel graphical method, Poem Emotion Trajectory System (PETS), is proposed to depict the flow of emotion in a poem. PETS is based on the construction of a weighted directed graph as a means to represent the emotion flow among the verses of a given poem. The weights represent the transition probability among the emotion states considered. The significant advantage is that a dominant path for each emotion category is identified. Emotion flow along verses is analyzed using a graph-based approach. This method, applied to each emotion category, generalizes the emotion flow in each emotion class. This PETS can be applied in poetry therapy and to enhance creative thinking and writing.

Automatic Authorship Detection Using Textual Patterns Extracted from Integrated Syntactic Graphs

We apply the integrated syntactic graph feature extraction methodology to the task of automatic authorship detection. This graph-based representation allows integrating different levels of language description into a single structure. We extract textual patterns based on features obtained from shortest path walks over integrated syntactic graphs and apply them to determine the authors of documents. On average, our method outperforms the state of the art approaches and gives consistently high results across different corpora, unlike existing methods. Our results show that our textual patterns are useful for the task of authorship attribution.