Week 17: 12/15/2015-12/22/2015

This past week I completed the first draft of our proposal for Tapia 2016. I also wrote the abstract for our poster proposal for Tapia. I discussed how we seek to explore expertise prediction for software developers. We compare three different machine learning algorithms that are used for our prediction. We have completed testing for the two-class boosted decision tree algorithm and will discuss the next two algorithms to use in future testing. With 85% accuracy we can conclude that we can predict software developer expertise using a two-class boosted decision tree machine learning algorithm.

I also finished work on the initial pass of the VISSOFT 2016 website. The website has been published and can be viewed at the following link:

http://vissoft16.ysu.edu/

Happy Holidays!

Weeks 15,16: 12/01/2015-12/15/2015

That last two weeks were filled with final projects and tests. On Sunday, December 13, 2015, I graduated!

Anyways...
Last week, I read and summarized "Tracking Students’ Cognitive Processes during Program Debugging – An Eye-Movement Approach" by Y. T. Lin, Member, IEEE, C. C. Wu, Y. C. Lin, T. Y. Hou, F. Y. Yang, and C. H. Chang. This paper discussed the differences between high performing(experts) and low performing(novices) students in debugging. 

Weeks 14,15,16: 11/25/2015-12/15/2015

The past few weeks have been very hectic with the end of the Fall semester. Last week was our finals week at YSU. However, I have continued to work on our CREU project.

I wrote the data section of our Tapia poster proposal submission and started to write the introduction. I read two papers, Modeling How Students Learn to Program by Chris Piech and Paulo Blikstein and Tracking Students' Cognitive Processes during Program Debugging - An Eye-movement Approach by Lin et al. I also completed my CREU mid-year report.

I have also been working on the VISSOFT 2016 website, because I am the website chair for this conference.

Week 14: 11/25/2015-12/01/2015

I read and created a summary Powerpoint for Modeling How Students Learn to Program by 
Chris Piech, Mehran Sahami, Daphne Koller, Stephen Cooper, Paulo Blikstein.
It was interesting. It talked about how to use machine learning to model how students progress 
through writing a program. This is interesting to me because I eventually want to pursue educational
technologies.

2

Weeks 12 & 13 11/11/2015-11/25/2015

In the past 2 weeks I have done several things. I attended a webinar on the Tapia conference. I have been looking in depth trying to understand the machine learning experiments ran regarding expertise, class recommender, and class recommender using pupil excitation. I am waiting to receive a copy of a fellow student's project to see what has all been done.

I have also gone to a meeting regarding the funding for our ACM-W group. We received a good amount from the university. We are also getting things in order for the Hour of Code celebration and our table for the organizational fair in January.

Week 13: 11/17/2015-11/25/2015

This past week I worked on the issues wrong with the new feature we were adding to iTrace. The feature allows us to gather eye-tracking data on different parts of the Eclipse Workbench.

We were having trouble with the Project Explorer part gobbling up gazes that should be assigned to a StyledText part. I believe the Eclipse UI is not set up the way we expected it to be or it could be an issue with the threading/architecture of iTrace.

That was all that I could deduce in the past week and I will continue to look into these issues in the coming weeks.
We are continuing on with a study using iTrace without this added feature. The data we collect during this study will add to our data analysis for CREU.

Week 12: 11/11/2015-11/17/2015

This past week I finished lab 4a from the Edx labs. In the lab, I learned how to construct and evaluate regression machine learning models using Azure ML and R.
More specifically, I trained and evaluated a nonlinear regression model which produces improved predictions of building energy efficiency.

I also spent some time setting up the SERESL lab for testing of a new feature we added for iTrace. The feature allows us to gather eye-tracking data on different parts of the Eclipse Workbench that was not available in iTrace before. Preliminary testing revealed issues in the new feature that will be addressed in the coming week.

Week 11: 11/4/2015-11/11/2015

This past week I worked on a template for our poster proposal for Tapia 2016. We will discuss a few preliminary machine learning experiments we performed on the data. We will discuss the data and its acquisition, the methodology for predicting expertise and difficulty, and the preliminary results. We will also have a small literature review and we will have an introduction that includes our hypotheses for data analysis.

Week 11: 11/4/2015-11/11/2015

This week I began to try and understand the machine learning experiments ran on the data sets we will use for the Tapia paper. I am trying to get a good grasp on what the different methods are. 

I started to get prepared for my meeting with the student government association to get funding for our ACM-W group. We hope to participate in the Hour of Code during the week of December 14.

I have also submitted the rest of my paperwork for graduate school. 

Week 9 &10 10/20/2015-11/4/2015

Over these two weeks, I completed lab 4a. I learned how to construct and evaluate regression machine learning models. Regression is a fundamental machine learning method. Regression enables you to predict values of a label variable given data from the past. Regression, like classification, is a supervised machine learning technique, wherein models are trained from labeled cases.  

I also completed  lab 4b in Edx. It demonstrated how to train and evaluate a classification model. Classification is fundamental in machine learning. Classification models enable you to predict classes or categories of a label value. 

I submitted the seed fund application for ACM-W. I also sent in the application for funding through YSU. 

Week 10: 10/27/2015-11/4/2015

This past week I worked on another Edx lab, Lab 4a. The goal of the lab is to learn more about using regression on data. I trained and evaluated a nonlinear regression model which produces improved predictions of building energy efficiency.

I also spent some time researching the topic of regression to get a better feel for how it would apply to our data.

I participated in the ACM-ICPC programming competition this past weekend. My team was able to solve 2 problems and placed 26th in the region.

Week 9: 10/20/2015-10/27/2015

This past week I completed another Edx Lab (4b). In the lab, I trained and evaluated a classification model. Classification is one of the fundamental machine learning methods used in data science. Classification models enable you to predict classes or categories of a label value. Classification algorithms can be two-class methods, where there are two possible categories, or multi-class methods. Like regression, classification is a supervised machine learning technique, wherein models are trained from labeled cases.
This lab is important, because we will perform classification on our ABB data.

I also spent some time figuring out what another classification experiment was doing in our SERESL Azure space. The student was performing multiple two-class bayes point algorithm on different eye-tracking categorical data from each task performed in the study. He also ran a two-class decision forest algorithm on task 2 in multiple different ways.

Week 8: 10/13/2015-10/20/2015

I continued to work on the Edx labs for Azure. I learned how to use R to manipulate and analyze data in Azure ML. I learned how to sample and quantize data in Azure. I also learned how to perform ‘data munging’.

I have also been working on debugging the code I wrote for our eye-tracking software, iTrace. The code will allow us to gather eye-tracking data on Stack Overflow documents.

Week 8: 10/13/2015-10/20/2015

I have continued working on the Azure labs. So far I have learned how to set up and score a model, how to upload data, how to combine multiple data sources. I have also learned how to use R and Python within Azure. 

Tomorrow, I take the GRE test. I began studying over the summer so I hope I do well. 

Week 7: 10/6/2015-10/13/2015

This week I continued working on the Edx Azure labs. I have start looking at how to program in python for Azure. I have been using http://www.learnpython.org/ to learn the basics. I have also finished the registration form for the YSU chapter of ACM-W.

Week 7: 10/6/2015-10/13/2015

This past week I ran two lab experiments in Azure. I was following the first two labs in Edx to get me started in learning how to use Azure. We will be using Azure to run machine learning algorithms and experiments on our ABB data, so it is important that we learn how to properly use it. The first lab involved a very basic introduction to a simple experiment flow. The second lab taught me how to upload data files to Azure and how to use the Join module to combine data from multiple sources.

Week 6: 9/29/2015-10/6/2015

This week I finished running the data through Itrace. I began it last week, however there was an error in the code that needed fixed before we could continue.
I also began reading up on how to Microsoft Azure. Azure is a cloud computing platform that we will use for data mining and prediction. I signed up for a Edx class that teaches how to use Azure.

Week 6: 9/29/2015-10/6/2015

This past week I helped Jessica run iTrace's fixation filter on the raw ABB data for our analysis. We went through each of about 20 participant folders and selected the 3 task XML files for the fixation filter to run on. It provided us with merged raw gazes based on an algorithm developed by Tobii for each of the tasks for each participant also in XML format. We had a few issues with a few files, because of their naming conventions, but we fixed those to get accurate data.

Week 5: 9/22/2015-9/29/2015

This week  I ran the data obtained from a study through fixation filters using iTrace

Week 5: 9/22/2015-9/29/2015

This past week, I ran our ABB study JabRef copies through srcML. I also spent time learning how to use Microsoft Azure to perform machine learning on our data. I ran a test experiment to get an idea of the work flow.

Week 4: 9/15/2015-9/22/2015

I spent the week working on mostly ACM-W stuff. I re-read the Conati paper to look better at the analysis.

Week 4: 9/15/2015-9/22/2015

This week I spent some time making sure the iTrace could correctly filter our raw gaze data into fixations in large batches. I also purchased an external hard drive, to be able to get the ABB data that we will be performing our predication analysis on.

Week 3: 9/8/2015-9/15/2015

This week I read Tracing Software Developers' Eye and Interactions for Change Tasks by Katja Kevic, Braden Walters, Timothy Shaffer, Bonita Sharif, David Shepherd, and Thomas Fritz.
I also sent in all the paperwork for CREU.

I also worked on writing the bylaws for our ACM-W group! I am meeting with student activities this week to get more guidelines for the group.

Week 3: 9/8/2015-9/15/2015

I read two papers this week. I read Inferring Visualization Task Properties, User Performance, and User Cognitive Abilities from Eye Gaze Data by Ben Steichen, Cristina Conati, and Giuseppe Carenini and Tracing Software Developers’ Eyes and Interactions for Change Tasks by Katja Kevic, Braden M. Walters, Timothy R. Shaffer, Bonita Sharif, David C. Shepherd, and Thomas Fritz.

I also filled out and submitted my CREU 2015-16 required documentation paperwork.

Week 2: 9/1/2015-9/8/2015

This week I finished reading and summarizing "Inferring Visualization Task Properties, User Performance, and User Cognitive Abilities from Eye Gaze Data" by Ben Steichen, Cristine Conati, Guiseppe Careinini.

I also began to read "Tracing Software Developers’ Eyes and Interactions for Change Tasks" by Katja Kevic, Braden M. Walters, Timothy R. Shaffer, Bonita Sharif, David C. Shepherd, Thomas Fritz. 

At our meeting we sat down and discussed what we plan to accomplish this semester. 

Week 2: 9/1/2015-9/8/2015

Dr. Sharif and I were in Bergamo, Italy for FSE 2015 from 9/1-9/5. I have attached some photos from our experience.
I had a very good time. Dr. Sharif and I presented a tool demo for iTrace, our eye-tracking plug-in for Eclipse (being developed in YSU's Software Engineering Research and Empirical Studies Lab). I met a lot of different faculty and students from across the world. I was able to listen to many different talks related to Software Engineering and talks related to the research we are doing with CREU.

Week 1: 8/25/2015-9/1/2015

I'm excited to be back for a new semester!

This week I began to read "Inferring Visualization Task Properties, User Performance, and User Cognitive Abilities from Eye Gaze Data" by Ben Steichen, Cristine Conati, Guiseppe Careinini. 

I also got the Emotiv Epoc EEG device, we had purchased in the summer, sent in to get it fixed.