Hi my name is Marcel Gentles and I worked with STEAM in AI to make a project

to make a tool for  young people teens and young adults who are looking to enter the workforce to analyze their social media presences and make sure that the their online presence and the kinds of content they’re putting up on there are only going to help them and not hurt them

so to achieve this

I worked with a mentor named Troy Kling from UC Santa Barbara he’s a math he’s getting a PhD in math there and he has really strong Computer Science Background so he helped me learn all about machine learning AI, what the differences are

and together

we created a sentiment analysis tool

so we pulled we interface with

Twitter’s API using a python Library

called Tweepy and throughout this whole process I went from starting out learning python that was the first couple months of this project was just me learning Python and understanding,

then next was understanding machine learning after that I actually began to work on coding scripts uh for example

first to pull tweets from Twitter to build a data set to work with then I learned about some different machine learning models and coded

I coded a training algorithm into them so I was able to code enough where I could train these machine learning models and extract features from the tweets and learn all this cool machine learning stuff and the end result was a tool where you could type in a username and specify some parameters like how many of the last how many of your most recent tweets would you like to look at would you like to look at your retweets as well and maybe your replies all different kinds of little details about your Twitter but

using those parameters and could return to you a list of your recent tweets however many you specified along with a link to each one so you can click it and then take action if it needs to be looked over or deleted or if you just want to see who’s interacted with it you can all do it from

my application so yeah we put that together and oh and then also it gives you a label and I what I didn’t do but I’m actually looking looking into in the future is uh redoing this project but instead of just a distinct label as positive or negative maybe do something where it can fall somewhere in between and that’s a machine learning thing called regression as opposed to classification putting things in two distinct classes or two or multiple distinct classes

so I learned a lot during this project

some of the some of the harder lessons were

that especially when you’re just busy and you have a lot of things going in your life you need to if you’re working on a big project like this you need to pace yourself and make sure to work chip away at the project little by little instead of trying to do it in little and huge chunks that take a short amount of time because that’ll leave you feeling stressed out it’ll leave you kind of forgetting what you’re doing in the time between your big Sprints

and beyond that another thing I learned is that

it’s really important to leave documentation in your code so you have to leave enough comments and things like that for you to understand so for you to fix problems but also for other people to understand what’s going on with your code

because the most powerful tool in this whole realm is collaboration and when you leave clear comments about what your code is doing other people can help you out and work with it just like they programmed it themselves

so I guess some of the final things I want to say

are that I had just a really great experience with this program and I was able to put it on my college applications and show them that I care about making an impact on the world and I’m dedicated to learning about these topics because it’s really something that I’ve discovered I’m passionate about and yeah so this is left me feeling like

I want to continue doing this in the future and I’m really looking forward to computer science

so thank you, DataEthics4All!!