SimMe

Document Type

Abstract

Publication Date

Fall 12-1-2024

Abstract

In an age where digital communication dominates, SimMe offers the possibility of creating a virtual version of oneself that can continue conversations, represent the user, and interact with others even when the user is unavailable. SimMe is a web-based application designed to achieve exactly that. By using personal message data and sophisticated AI techniques, SimMe creates a highly personalized virtual self, capable of engaging in realistic, human-like interactions. SimMe leverages Natural Language Processing (NLP) and Large Language Models (LLMs) to analyze message histories, learning the user's unique conversational patterns, tone, and style. This virtual self can then be deployed in various scenarios—whether responding to friends in a message queue, scheduling tasks when the user is busy, or maintaining an ongoing conversation when the user is unavailable. The project's ultimate goal is to enable users to engage with SimMe through a web interface, where the virtual self generates responses tailored to each specific target, based on prior interactions. This can be useful in several contexts: for instance, managing conversations that require immediate responses, acting as a stand-in during messaging downtime, or even as a personal social experiment. Beyond its practical applications, SimMe also serves as a unique experiment in AI-driven social interaction and virtual self-representation. Ethical considerations are central to the project, with strict measures in place to ensure data privacy and user consent for every conversation analyzed. In short, SimMe opens new possibilities for personalized AI interaction, allowing users to seamlessly extend their digital presence.

Comments

Completed as part of the Computer Science Senior Capstone Project.

This document is currently not available here.

Share

COinS