Text Generation with GANs
Project Abstract
This project is undertaken with a focus on exploring the frontiers of automated text generation through the lens of Generative Adversarial Networks (GANs). Aimed at creating short poems, stories, and short text based on prompts, this project seeks to uncover the synergies between technology and linguistic creativity. By leveraging the capabilities of GANs, The project aims to generate text that demonstrates coherence, inventiveness, and contextual relevance.The methodology for this project revolves around training the GANs using a dataset rich in narratives and textual snippets. This foundational step enables the network to assimilate and mirror the complexities and subtleties inherent in human language. A evaluation process will follow to examine the quality, originality, and applicability of the generated text, emphasising its alignment with accepted norms of authenticity and practical usability.This project aspires to culminate in the development of a GAN model that seamlessly combines technical skill with the creative spirit of storytelling and text expansion. This model is envisioned as a platform that heralds a nuanced approach to automated text generation, fostering innovation and enhancing textual creativity and convenience.
Keywords: Artificial Intelligence (AI), Data Science, Neural Networks
Conference Details
Session: Poster Session A at Poster Stand 66
Location: Sir Stanley Clarke Auditorium at Tuesday 7th 13:30 – 17:00
Markers: Benjamin Mora, Fabio Caraffini
Course: BSc Software Engineering, 3rd Year
Future Plans: I’m undecided