HeaderSIS.jpg

Difference between revisions of "IS480 Team wiki: 2017T1 Team Atoms Final"

From IS480
Jump to navigation Jump to search
Line 223: Line 223:
  
 
====Recap====
 
====Recap====
A summary from our previous complexities achieved : https://wiki.smu.edu.sg/is480/IS480_Team_wiki%3A_2017T1_Team_Atoms_MidTerm#Technical_Complexity  
+
A summary from our previous complexities achieved -
 +
https://wiki.smu.edu.sg/is480/IS480_Team_wiki%3A_2017T1_Team_Atoms_MidTerm#Technical_Complexity  
 
<center>
 
<center>
 
{| class="wikitable" style="font-family:Roboto; background-color:#FFFFFF; width: 1000px;" align="center"
 
{| class="wikitable" style="font-family:Roboto; background-color:#FFFFFF; width: 1000px;" align="center"

Revision as of 02:25, 18 November 2017

Atoms Logo.png


Team Atom Icon Home.png

HOME

  Team Atom Icon Aboutus.png

ABOUT US

  Team Atom Icon Projectoverview.png

PROJECT OVERVIEW

  Team Atom Icon Projectmanagement.png

PROJECT MANAGEMENT

  Team Atom Icon Documents.png

DOCUMENTATION



Project Progress Summary


Atoms finals dashboard.png

Atoms finals slideicon.png Atoms finals application.png Atoms finals videoicon.png Atoms finals postericon.png

User: demo | Password: demopassword123

Project Management

Project Status

Atoms finals projectstatus.png

Project Schedule (Plan vs. Actual)

There were afew changes made to the project schedule and scope as the team decided to push themselves and added new refinements to the project. In addition, the sponsor has requested for an additional technical documentation due to their intentions to open source the project and submit a software research publication around the application. The changes made did not affect the timely completion of the project. We have also brought the project handover forward to allow more time for the sponsor the take over the project and provide sufficient time for support.

Planned Project Schedule

Atoms Schedule V7.png

Actual Project Schedule

Atoms finals actualschedule.png

Project Metrics


Atoms schedule metric description.png
Atoms schedule metric.png
Schedule Metric Breakdown: https://www.dropbox.com/s/tmpczn3wr520nqf/Project_schedule%20.xlsx?dl=0


Atoms bug metric description.png


Atoms bug metric.jpg


Atoms bug score.png

Change Management

Below is the change log for Iteration 11 to 14 (after Mid-Terms):

Change Log
Iteration Date Type Change Request Rationale Feasibility Outcome Priority Status of Request Issued By
11 10/10/2017 Scope Add interactive GIF tutorial (For first time users) Supervisor recommended including a interactive tutorial guide for first time users to allow them to easier learn how to use the system Fits into schedule without any expected delay as a consequence Accepted Low Closed Supervisor
12 16/10/2017 Scope Change denogram implementation for clustering (Distance cut off) Sponsor requested for a better way to view denogram and allow users to utilise it more effectively through distance cut off Fits into schedule without any expected delay as a consequence Accepted High Closed Sponsor
12 16/10/2017 Scope Add interactive association visualisation Team decided to make visualisaiton for association link graph interactive as it adds more value to the visualisation and project Fits into schedule without any expected delay as a consequence Accepted Low Closed Team
12 18/10/2017 Scope Deployment documentation instructions Team decided to create an additional set of deployment documentation to facilitate project handover Fits into schedule without any expected delay as a consequence Accepted Low Closed Team
12 20/10/2017 Scope Technical documentation of system Sponsor request for technical documentation as they will like to open source the project and submit for software publication Fits into schedule without any expected delay as a consequence Accepted High Closed Sponsor
13 30/10/2017 Schedule Change implementation for association rules Lab4 release have to be delayed by 1 week due to changes in association computation/ algorithm Fits into schedule without any consequence Accepted High Closed Sponsor

Project Risks


Atoms Risk category.png


Existing & Potential Risk

Currently there are no outstanding risk. All identified risk and challenges have been addressed. However, from the period of Mid-Terms till before Finals , we have faced and resolved concerns arising from 1) Client Management Risk as described below:


Risks & Challenges Faced

S/N Risk Type Risk Event Likelihood Impact Category Mitigation Strategy
1 Client Management Risk Team Atoms will face modifications to be made within a short period of time before every lab release. As the team will have to provide a highly customized and similar lab session experience with the existing lab exercises used in class. In addition, the team will also have to provide a new highly customized system user guide with every lab release. High High A Team will actively engage sponsor as early as possible before each lab release to provide sufficient time for changes required. In addition, the team will also dedicate time to specifically prepare for each lab release.

Technical Complexity

System Architecture

Atoms system architecture.PNG Atoms system architecture2.PNG

Recap

A summary from our previous complexities achieved - https://wiki.smu.edu.sg/is480/IS480_Team_wiki%3A_2017T1_Team_Atoms_MidTerm#Technical_Complexity

S/N Technical Complexity Risk Event
1 Canvas Graph Traversing Algorithm Team Atoms designed a graph traversing algorithm to handle all the possible combinations the user could draw in the canvas
2 Concurrency issue with Django + Matplotlib Team Atoms handled concurrency issues causing visualization charts to overlap each other (which becomes unreadable)
3 Ensemble Algorithm Team Atoms wrote our own Ensemble algorithm where we implemented our own Voting Classifier.

The following are new technical challenges faced by the team from Mid-Terms till Finals:

Frontend

1. Interactive Association Graph

In order to provide a better visualization when rule based mining is applied, we used d3.js to visualize the association rules on an interactive graph instead of a static image visualization. This will enable the users to move the nodes around to get a clearer view of the different rules generated. In addition, we were also able to achieve a high level of customization for the visualization by configuring the line color, node color and even node size based on its confidence and lift levels and items length.

Atoms finals interactive graph.png

As the team was new to the D3 visualization library, there was a steep learning curve for us. The library was challenging to understand and grasp effectively in a short period of time. In addition, changes also have to be made on the backend where we had to modify the structure of the dataset in order to make it compatible with D3.



Backend

2. XGBoost Classification Implementation

What is XGBoost? XGBoost is short for eXtreme gradient boosting. It is a library designed and optimized for boosted tree algorithms. It's main goal is to push the extreme of the computation limits of machines to provide a scalable, portable and accurate for large scale tree boosting. The term “Gradient Boosting” is proposed in the paper Greedy Function Approximation: A Gradient Boosting Machine, by Friedman.

Implementation
We implemented the XGBoost Classification algorithm using the Python API (http://xgboost.readthedocs.io/en/latest/python/python_api.html) and plugged it into our functionalities. Since the library also provides a Scikit-Learn Wrapper, it should work natively with our features:

  • Generate Accuracy Report
  • Generate Graph (Using Graphviz)
  • Generate Confusion Matrix
  • Ensemble
  • Predict

Complications
However, complications arose when we used a dataset with columns that contained spaces -- the Graphviz library did not handle spaces properly when performing text splitting, causing the plot to crash.

Atoms finals xgboost.png

Outcome: Rules not split correctly - Strings broken up

Atoms finals xgboost1.png

Fixes Attempt #1: Initially, we wanted to override the library to write a custom split but it can cause future implications when the new version releases etc. So we replaced the spaces between the strings with an underscore. Everything seems to work fine until we connected the XGBoost to an Ensemble and it did not work because other classifiers were having different feature names (without spaces)!

Atoms finals xgboost2.png


Attempt #2: Instead of changing the feature names permanently, we only modified it when it went into the graphing library and replaced it back with the original (into the XGBoost Model attribute) after the export graph was complete.

Atoms finals xgboost3.png

Results:

Atoms finals xgboost4.png



Quality of Product

Recap

A summary from our previous measures in place to ensure that we deliver a quality product-

https://wiki.smu.edu.sg/is480/IS480_Team_wiki%3A_2017T1_Team_Atoms_MidTerm#Technical_Complexity 
S/N Quality Measures Description
1 Canvas Graph Traversing Algorithm Team Atoms designed a graph traversing algorithm to handle all the possible combinations the user could draw in the canvas
2 Concurrency issue with Django + Matplotlib Team Atoms handled concurrency issues causing visualization charts to overlap each other (which becomes unreadable)
3 Ensemble Algorithm Team Atoms wrote our own Ensemble algorithm where we implemented our own Voting Classifier.

The following are new technical challenges faced by the team from Mid-Terms till Finals:

1. Deployment Script

Manual deployment can lead to multiple human error. Hence, we have created a deployment shell script that partially automates the process of the deployment of our web application. The steps that are automated includes:

1. Stopping/Starting of system services running our web server and our web application
2. Downloading of new source code from git repository
3. Changing file system permission of directories and files
4. Execution of Django specific deployment command

With a frequent deployment rate (every iteration - 2 weeks) the chances of error due to manual deployment is much higher. Hence with the deployment script we will be able to reduce such errors.


2. Bench marking for Visualization

This is a scatter matrix plot for the famous Iris data set. It involves plotting every column against one another. Therefore, the computation is (# of rows) x (# of columns)2
Complexity: Atoms midterm benchmarkingcomplexity.png

Atoms midterm benchmarking iris.png

Problem

  • This would cause datasets which a large number of columns to take a considerable amount of time, consuming resources for a single user
  • If many users execute this chart at the same time, it would result in a very long response time


Solution

  • To accurately measure how much time it would take for different dimensions of data sets, we generated datasets of different columns and rows and ran each charting function to see how much time it took.
Atoms midterm benchmarking timing.png

Findings

  • From the benchmark tests, we found out that the number of rows did not affect the performance as much as the number of columns
  • From our findings, we implemented validations in place to disallow users to select too many columns (>10) for scatter matrix



3. Secure API - System security

  • All backend APIs require user login to prevent unauthorized direct API calls
  • For each API request to modify files, there is an implementation to verify if the file belongs to the user before the operation.
  • If there is an unauthorized API request to the system, an appropriate error message will be displayed and the request will also be logged for investigation
Atoms midterm secureapi.png


4. Google Analytics tracking implementation

Google analytics tool can help our sponsor understand how students are interacting with our KDD Labs website, where they’re coming from and how often they visit, what parts of the site are capturing their attention and what parts aren’t sparking interest.

Atoms midterm googleanalytics.png


We also keep tracking user's’ behaviour and count number of events triggered in our system. This will allow us to keep a close monitor on functionalities being utilized in our system and assist us with tracking abnormal behavior. Furthermore, this will also be a useful tracking tool for the teaching team to understand the students usage behaviour. There are a total of 32 different activities that have been tracked since the website was announced to the students on the 7th Sep 17. The graph below shows the top 10 activities in our website.

Atoms midterm googleanalytics top10.png


5. System logger

Our system will consistently monitor and log down critical user actions and the problems the user encounters. This will help us to automatically track errors made in the system which will be used for our internal feedback when users utilize the KDD Labs system. We will then analyse these errors further and derive the root cause of such errors to try improve on our system if possible.

Atoms midterm systemlogger.png


Furthermore, to make the logging files easy to locate, we have created a logger that will rotate the logging file twice a day. The file name will be changed to the date and time when the file was last modified.

Atoms midterm systemlogger rotation.png


Intermediate Deliverables

Topic of Interest Link
Project Management Project Schedule
Minutes
Metrics
Risk Management
Change Management
Project Overview Project Overview
Team's Motivation
Project Scope
Project Documentation Diagrams
Technologies Implemented
Low & Mid Fidelity Prototypes
Testing Testing Documentation


Deployment


To view application, visit
Test server: http://kddlabs.com
Username: demo
Password: demopassword123


Note: This public server is currently being utilized for Live Usage by the IS424 Data Mining and Business Analytics students for their labs and project completion.
Production server: https://kddlabs.cn/

Testing

Atoms internaltesting.PNG Atoms UAT1.PNG Atoms UAT2.PNG Atoms Liveusage.PNG

Internal Testing

We engage in comprehensive manual testing in every iteration. The developers will conduct individual testing before committing their codes on our shared repository, GitHub. We believe in testing the application manually at this level because tests can be specially adjusted to cater to changes in the application, both on the front and back end. Furthermore, manual testing brings about the human factor, allowing us to better discover problems that might surface during real usage due to natural human behavior.

Once the developers have fixed the bugs, the fixed set of codes will be merged and integrated with the other functionalities. Subsequently, the integrated code is then deployed on the test server and the lead quality assurance will run a final check against the set of test cases created. This helps to ensure that the deployed application works with no major incidents.

The team's lead quality assurance then performs regression testing on the test server where previous functionalities developed are tested again. This helps to ensure that existing functionalities in the application are not affected by the integration. Once bugs have been identified, the lead quality assurance will then update the bug-tracking Excel sheet and notify the relevant developers of the issues and the corresponding priority level.


The team’s list of test cases can be found on our private repository here.

User Acceptance Test 1 & 2

Team Atoms has conducted 2 user tests which allowed us to better manage sponsor expectations as well as improve on usability of our application interface.

Atoms midterms uat summary.png

For more detailed version of Team Atoms user acceptance test results, access it here:

Atoms UAT1.PNG     Atoms UAT2.PNG


Live Usage

Through our live usage and roll out, IS424 students were able to complete their take home lab assignments on our system. Thus we were able to gather feedback from our end users about KDD Labs system directly. In addition, we were also able to compare the user experience with the existing alternative used in class (SAS EM). From our feedback, we have received positive response about the KDD Labs system stating that the students were able to complete their in class lab exercise. Not only that, they also found that the KDD Labs system was easier to use as compared to SAS EM.


Lab1

Release Date: 01 Sep 2017, Friday
Duration: 2-3 hours per user
Number of Users(s): 45 
Lab1 User Guide: Created instructions can be found here
Lab1 Feedback Results: Live user feedback can be found here

Atoms lab1feedback Q1.png
Atoms lab1feedback Q4.png
Lab2

Release Date: 15 Sep 2017, Friday
Duration: 2-3 hours per user
Number of Users(s): 45 
Lab2 User Guide: Created instructions can be found here
Lab2 Feedback Results: Live user feedback can be found here

Atoms Lab2 feedback.PNG
Atoms Lab2 feedback1.PNG
Lab3

Release Date: 16 Oct 2017, Monday
Duration: 2-3 hours per user
Number of Users(s): 45 
Lab3 User Guide: Created instructions can be found here
Lab3 Feedback Results: Live user feedback can be found here

Atoms Lab3 feedback.PNG
Atoms Lab3 feedback1.PNG
Lab4

Release Date: 6 Nov 2017, Monday
Duration: 2-3 hours per user
Number of Users(s): 45 

Reflection

Team Reflection

This journey has proven to be an enriching learning experience for Team Atoms. The project had many new learning points for the team as it was highly technical- we had to understand and grasp the concepts of the data mining process and algorithms within a short period of time. In addition, we also learnt the importance of good stakeholder management which allows us to better react to unforeseen circumstances. Through an active team participation and communication we were able to mitigate existing issues and deliver a quality project on time.

Sponsors' Testimonial


Atoms midterms sponsor.PNG

"Team ATOMS is a capable and sincere team, that has done very well in the course of the project. KDD Labs project is very challenging and in particular requires a very diverse set of technical skills. ATOMS have made substantial efforts in acquiring new skills and integrating them to deliver a quality product in a timely manner. They have stoically faced the technical issues and challenging feature/change requests, and demonstrated an excellent work ethic in delivering on their targets. Discussions with them have been thought provoking and rewarding, and have significantly contributed towards improving the product quality" - Sponsor, Doyen Sahoo

Individual Reflections

Atoms midterms reflections.PNG