Shelby Hall Graduate Research Forum Posters

Files

Download

Download Full Text (512 KB)

Description

Stream of consciousness writing has a long history, including novelists James Joyce and Virginia Woolf. However, there has been little work done in automated and semi-automated analysis of such writing, which is the focus of this work. We plan to divide real streams of consciousness writing into distinct topical units and then capture different momentary meaningful topics from these units. By doing this, researchers and readers could gain a more nuanced understanding of the narrative structure and thematic elements. In addition, it would also support applications in fields like psychology and linguistics, where understanding thought processes and narrative structures is crucial. To accomplish this task, we conduct an initial manual segmentation and then refine the segmentation using automated tools. First, we perform topical segmentation, a natural language processing approach that divides text into semantically coherent sections or topics. Next, we use topic modeling to combine the segments via methods such LDA or deep learning. This results in a macro-level representation of the evolution topics in stream of conscious writings along with a listing of finer grain segments. To test out approach, we use publically available collections of writings, including children stream writing, collections of graduate admission essays, along with examples from stream consciousness novels. Initial qualitative results indicate our model effectively identifies narrative flow sections where the focus shifts significantly along topical boundaries. Our future work will include additional qualitative and quantitative analysis, along with examining applications in the fields of musicals with linguistics stream expression.

Publication Date

3-2025

Department

Computer Science

City

Mobile

Disciplines

Data Science | Graphics and Human Computer Interfaces | Other Computer Sciences | Programming Languages and Compilers

Topical Text Segmentation for Stream of Consciousness Writing

Share

COinS