2016 Computer Science Senior Design Projects

Student Report Title Student Report Title

Some links below may require MS Powerpoint 2007 or Adobe Reader
Click on a student picture for a larger view.

Jeffrey T. Cohen

Advisor: Prof. Matthew Anderson

SoundByte: An iOS Application to Enhance Music Discovery

This thesis examines the process from conceiving an idea for a mobile application to building an iOS application that consumers will theoretically use, to App Store inception. The mobile application, Sound- Byte, is positioned to serve as a solution to optimize music discovery efficiency. The application solves an issue that consumers face on a daily basis: Not having a proficient way to find new music that they enjoy.

Poster Link    Report Link    Presentation Link   

Andrew Colello

Advisor: Prof. Valerie Barr

Image Blob Detection: A Machine Learning Approach

The field of computer vision has developed several methods of detecting contiguous circular pixel ar- eas, or blobs. While both blob detection and computer vision are both still in their infancy, they are already being used in a variety of applications with significant successes. This study examined the effectiveness of classification-based machine learning algorithims on blobs and compared several models with a baseline to measure their performance. The system’s objective was to take an image of a golf course as input and output the location of a golf ball in the image, if one is detected. The system took a raw image file as input, scanned it for blobs using image processing software, and then classified each blob as positive or negative as output. Results showed that several classification models, especially tree-based models such as Random Forest, could be used to classify blobs with statistically significant success.

Poster Link    Report Link    Presentation Link   

James E. Curbow

Advisor: Prof. Matthew Anderson

Blending Two Automatic Playlist Generation Algorithms

We blend two existing automatic playlist generation algorithms. One algorithm is built to smoothly transition between a start song and an end song (Start-End). The other infers song similarity based on adjacent occurrences in expertly authored streams (EAS). First, we seek to establish the effectiveness of the Start-End algorithm using the EAS algorithm to determine song similarity, then we propose two playlist generation algorithms of our own: the Unbiased Random Walk (URW) and the Biased Random Walk (BRW). Like the Start-End algorithm, both the URW algorithm and BRW algorithm transition between a start song and an end song; however, issues inherent to the Start-End algorithm lead us to believe that our algorithms may create playlists with smoother transitions between songs.

Poster Link    Report Link    Presentation Link   

Duri Abdurahman Duri

Advisor: Prof. Chris Fernandes

The Effects of Disruptive Communication on the Engagement of Users in the Collaborative Consumption Model

Collaborative consumption is a new term that has recently gained popularity. It describes the recent trend of renting or trading of products and services through apps and common day electronics. Uber and Airbnb are famous companies that are leading this trend and have created a sense of ’sharing’ within the economy. By building apps and creating a connection between people who share common interests, many companies have set out to become the ’Uber of X’ with the hopes of immense monetary returns. The popularity of these services has created competition to get more people using one service opposed to another. This research came from an intuition that the way users communicate with these apps affects how engaged users are with the app. Therefore, this project entails the creation of an online app that allows people to offer money in exchange for errands. It monitored the activity of users relative to how they are notified of new errand opportunities, primarily looking into text and email messaging relative to a user not receiving notifications at all. Data was collected for a seven-week period and analyzed for patterns in usage. Findings from this project report that notification type is indeed correlated with user engagement. Finds from this research also found that users who receive notifications through text messaging were the most engaged.

Poster Link    Report Link    Presentation Link   

Logan Fine

Advisor: Prof. Aaron Cass

The Presence of Timers and their Impact on Team Communications During High-Stress Scenarios

This research focused on the impact of a timer on the team dynamic in high-stress scenarios. Partici- pants were asked to defuse a virtual bomb while communicating via headset audio. Only one participant could see and interact with the bomb, while the other read and interpreted instructions from the bomb’s defusal manual. Results seem to suggest that groups where both participants can see a countdown timer for the bomb perform better than groups where only the bomb defuser could see it, due to higher rates of task completion, completion speed, and more effective communication. There is not enough data to be statistically significant.

Poster Link    Report Link    Presentation Link   

Joshua C. Loew

Advisor: Prof. Matthew Anderson

Examining the Viability of MINIX 3 as a Consumer Operating System

The developers of the MINIX 3 operating system (OS) believe that a computer should work like a television set. You should be able to purchase one, turn it on, and have it work flawlessly for the next ten years [6]. MINIX 3 is a free and open-source microkernel-based operating system. MINIX 3 is still in development, but it is capable of running on x86 and ARM processor architectures. Such processors can be found in computers such as embedded systems, mobile phones, and laptop computers. As a light and simple operating system, MINIX 3 could take the place of the software that many people use every day. As of now, MINIX 3 is not particularly useful to a non-computer scientist. Most interactions with MINIX 3 are done through a command-line interface or an obsolete window manager. Moreover, its tools require some low-level experience with UNIX-like systems to use. This project will examine the viability of MINIX 3 from a performance standpoint to determine whether or not it is relevant to a non-computer scientist. Furthermore, this project attempts to measure how a microkernel-based operating system performs against a traditional monolithic kernel-based OS.

Poster Link    Report Link    Presentation Link   

Anton Morozov

Advisor: Profs. Chris Fernandes (CS) and Steve Horton (Biology)

Automated Data Analysis Pipeline for Gene Expression Studies

In any particular tissue cell type, only some genes of an organism’s genome are active (expressed) at any given time. Identifying these active genes can help us understand what defines the cells identity. This is especially important in brain tissue, where neurons of different types have to interact in particular ways. Existing computational tools for gene expression analysis cannot effectively analyze data from poorly studied genomes, such as that of Aplysia, a model organism used in many neurobiological studies. To address this challenge and to find active genes that are associated with neuronal types, I have developed a computer pipeline suitable for processing both RNA-Seq and microarray gene expression data. The pipeline run requires little input from the user, and the results are presented via a web-based interface that accesses the projects database. Using this pipeline, I compared gene expression data derived from Aplysia motor and sensory neurons and found 185 genes and 24 gene pathways with significantly different steady-state levels. Many of these genes code for proteins that participate in cell interactions and signaling and reflect the functionality of the corresponding neurons. The developed pipeline can be used in a wide variety of projects dealing with gene expression analysis.

Poster Link    Report Link    Presentation Link   

Eric Rose

Advisor: Prof. Aaron Cass

Approaching Humans for Help: A Study of Human-Robot Proxemics

In order for a robot to be effective when interacting with a person, it is important for the robot to chose the correct person. Consider an example where a robot is trying to perform a task but it isn’t capable of doing a subtask, like going up a flight of stairs. In this case, the robot would need to ask a person for help with the elevator, in a socially appropriate way. We have conducted an experiment to determine who would be the best candidate to approach in a situation like this. Should the robot choose to approach someone who is very close, with the risk that the person may have already committed to passing? Or someone who is further away, which could result in the person not noticing the robot at all. Our hypothesis is there is some optimal distance that is not too close nor too far away. We tried approaching from different distances to see which distance led to the most successful interactions. The result provide guidance for developers of autonomous robots who need a robot to approach someone for help.

Poster Link    Report Link    Presentation Link   

Rodolfo Santana

Advisor: Profs. Chris Fernandes, Nick Webb

Power Outage Prediction via Twitter

There has been numerous amount of research done where researchers have mined social media in order to predict a variety of economic, social, and health related phenomena[2]. What I hope to accomplish is fine- grain prediction of outages across the US by monitoring tweets that either report or hint of one occurring. Data mining Twitter for information and using machine leaning methods to attain valuable data, I was able to evaluate roughly 5,000 tweets collected from December, 2015 though February, 2016. The system created showed that I am able to detect actual outages occurring at their precise location. There is still work to be done to be able to predict outages in a much more granular level, but as of now, some level of prediction was attained.The system plots cascading-failures as they unfold in real-time, which helps as a predictor of outages.

Poster Link    Report Link    Presentation Link   

Lily Steinberg

Advisor: Prof. Aaron Cass

On the Importance of Robot Faces for Human-Robot Interaction

Researchers have been doing experiments in the field of Human Robot Interaction for a while now. When dealing with their involvement, most researchers have been trying to maximally increase human comfort around robots by changing different factors and asking human subjects how they feel after. However, comfort and willingness to do a task are not the same thing. Throughout my research, I wanted to figure out if changing different factors about a robot’s physical appearance will increase human involvement in finishing a task and achieving a set goal. I tested their willingness by approaching people with a robot with a face and without a face. Rather than increase human comfort, I wanted to see if one would lead to a more successful interaction between human and robot.

Poster Link    Report Link    Presentation Link   

Computer Science Senior Design Projects from previous years:
2015 2014 2013 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003

All rights reserved - Union College, Schenectady, NY 12308
Last edit 26May2017
Address questions or comments to freyd@union.edu