Support Local Journalism


Subscribe


Like everything else in the COVID-19 era, Central Washington University’s Symposium and University Research and Creative Expression — also known as SOURCE —was streamed online all week, featuring a wide range of academic presentations.

Much like a normal year, everything from art to science to business to law was well represented in the student projects, which were uploaded to a platform where virtual attendees could comment and ask questions.

Andrew Dunn, from the school of graduate studies, used his SOURCE project to dive deeper into the world of artificial intelligence, one of his favorite topics.

Dunn graduated last spring from CWU with a bachelor’s of computer science. He’s currently researching data science and machine learning.

“I think they are really cool and you can do some crazy stuff with it,” Dunn said during his presentation.

For SOURCE, Dunn used a combination of Natural Language Processing and Machine Learning to train a neural network to analyze President Donald Trump’s tweets to determine whether they have a positive sentiment or a negative sentiment.

After a feeding the computer 1.5 million tweets that have been human labeled as positive or negative, Dunn had the neural network analyze some of Trump’s tweets.

The first example from the president was “I, for one, am glad to see he is back, and well!” Dunn’s model predicted it was a positive tweet with 74.12% confidence.

The next example from the president read “She was thrown off The View like a dog, Zero T.V. Personas. Now Wallace is a 3rd rate lapdog for Fake News MSDNC (Concast). Doesn’t have what it takes!” Dunn’s model predicted this was a negative with 83.63% confidence.

“It works, but it’s not perfect though,” Dunn said during his presentation. “The overall testing accuracy is around 80 percent. … One of the more exciting things we plan on doing is integrating sarcasm detection, so that’s coming in the future.”

AUTOMATED REPTILE HABITAT

Electronics engineering technology major Libby Wittman used her project to automate some of the time-consuming tasks necessary to take care of her reptilian pets.

Whitman has been keeping reptiles for eight years, and while she enjoyed it, she said it can often be overwhelming.

“It takes about a half hour daily just for misting,” she said during her presentation.

Her project automated the lighting, temperature and humidity of the reptile’s enclosure. She used a heat mat controlled by feed back from a temperature sensor, while the misting system was controlled by a humidity sensor. A lux sensor detected changes in ambient light in the room and turns on a light if the room is too dark, all pre-programmed to be set within daytime hours. The whole system is controlled by a small computer called a Raspberry Pi.

“This entire project is just a proof of concept that this can be done,” Wittman said during her presentation. She noted changes she might make in a future design would be a better pump designed specifically for reptile use, better nozzles that help create a fog instead of drops of water, and a cleaner graphical user interface for the software.

“This will improve the lives of just not the keeper, but the reptile as well,” she said.

PARKING ON CAMPUS

In life before COVID-19, parking on Central’s campus has always been a valuable resource. Paul McCafferty, Drake Wald, Corey Johnson, Kevin Bertelsen, Camilo Jacomet and Hailey Lawton — self-nicknamed “Hailey and the Boys” — developed an app to help find that resource.

“The No. 1 problem we hear from our peers about attending Central Washington University is the lack of parking on campus,” one of the group members said during the online presentation. “Our goal was to develop a mobile application that students, faculty and anyone else could use in order to quickly and efficiently find parking.”

An early problem the group encountered was the object recognition functionality — in other words, is there a car in the parking spot or not? The group said the learning curve for developing machine learning to detect objects was steep, and funding to mount cameras that would cover the lots required was going to be hard to come by.

The group designed the app to show the all the lot locations on CWU’s campus, which included filters for staff lots, student lots or free lots. It also informed the user of the number of available parking spots once they click on a specific lot.

While the group was able to create a functional user interface for both iOS and Android, they were unable to get the object detection and recognition working, but said it is a goal for the future and maybe another senior project.

Comments

Keep it Clean. Please avoid obscene, vulgar, lewd, racist or sexually-oriented language.
PLEASE TURN OFF YOUR CAPS LOCK.
Don't Threaten. Threats of harming another person will not be tolerated.
Be Truthful. Don't knowingly lie about anyone or anything.
Be Nice. No racism, sexism or any sort of -ism that is degrading to another person.
Be Proactive. Use the 'Report' link on each comment to let us know of abusive posts.
Share with Us. We'd love to hear eyewitness accounts, the history behind an article.