Computer Science

Luis Ayerdi-Morales, layerdim@ttu.edu
Texas Tech University, with Dr. Ziwen Pan

Evaluating the Security of Quantum Key Distribution Under Intercept-Resend Attacks: A Simulation-Based Study Using BB84
Quantum Key Distribution (QKD) offers a theoretically secure method for transmitting encryption keys using the principles of quantum mechanics. This research investigates how effectively the BB84 protocol can detect eavesdropping attemptsusing a basic intercept-resend attack model. We hypothesize that even simple eavesdropping strategies will introduce measurable errors in the communication channel, specifically increasing the Quantum Bit Error Rate (QBER), allowing thecommunicating parties to detect an eavesdropper.  To test this hypothesis, we implemented a simulation of the BB84 protocol in Python, modeling Alice and Bob exchanging qubits with and without the presence of an eavesdropper (Eve). Eve intercepts a varying percentage of the transmitted qubits, measures them in a random basis, and sends them to Bob, mimicking a realistic intercept-resend strategy. We measured QBER after basis reconciliation and plotted its behavior against different levels of eavesdropping.  Results showed a clear correlation between eavesdropping intensity and QBER, with QBER exceeding the lower bound standard security threshold of 11% when more than half the qubits were intercepted. This confirms BB84's built-inresilience against basic quantum attacks and its capacity to detect compromised transmissions.  This work demonstrates how QKD can maintain security through fundamental quantum principles and provides afoundation for exploring more advanced attacks. Future work may also include modeling physical qubit noise, extending simulations to other QKD protocols like E91 and weighing other possible applications of BB84 protocol.
 

Keith Howerton, khowerton@mail.snu.edu
Southern Nazarene University, with Dr. Rob Gering

Smart Speakers, Smart Ads: A Study of Voice Inputs and Advertising Algorithms
As voice assistants become ubiquitous in modem households, they raise significant privacy concerns about potential eavesdropping and targeted advertising. The study will help identify correlations between specific voice inputs on Google Home assistants and the appearance of related advertisements on Google services such as YouTube and Google Search engine. Three Google Home devices were used in the experimental design. One device served as a baseline of theexperiment and did not receive any voice inputs. The other two devices were tested with specific voice commands towardstravel and health to assess whether voice interactions influence targeted advertising. The data were manually collected onthree computers for four weeks. The first advertisements shown were recorded, including dates and URLs, and compiled inan Excel database. These records were then analyzed using a custom Python-based tool to detect patterns and potential correlations between voice inputs and ad content. Preliminary results suggest that there is no strong evidence that there is a correlation between the specific voice inputs devices and targeted ads. Future research could further examine theconditions under which voice data may be used for advertising purposes and explore more advanced privacy protections for consumers.
 

Rosa Ramirez, rosa.ramirez@csupueblo.edu
Colorado State University Pueblo, with Dr. Katie Brown

The Capabilities of Translation: Human vs. Machine
This presentation highlights the strengths and limitations of human translators and machine translators by comparing two English-to-Spanish translated text using both methods. The objective is to evaluate the accuracy, cultural presentation, and linguistic quality of the translation produced by Artificial Intelligence (AI) tools like ChatGPT and Microsoft Translator against those created by human translators. This research is significant because it addresses the growing reliance on AI in professional settings, where a more personal connection is needed to connect with people/readers. Two original English text documents titled Savages or Sophisticated and How Far did the Arawak Travel, were translated manually by the author and then through AI tools under control conditions. Four Spanish versions of the first text were generated using different prompts in ChatGPT and Microsoft Translator; the second document was translated once through each method for comparative purposes. While AI provides great speed and convenience in translation, findings show that it lacks the capabilities of preserving cultural elements, emotional tones, and contextual meaning that human translators naturally take into consideration. In conclusion, the findings suggest that despite the growth of machine translation, human translators remain essential for producing cultural sensitivity and emotional text that resonate with people on a deeper level, especially in academic, historical, or literary contexts.