Data Analysis and Taxonomy Redesign

PROJECT OVERVIEW

Goal: To solve Epic Games’ Unreal Engine 4 documentation challenges of discoverability and usability, causing confusion and frustration in users due to non-standard tagging vocabulary.
Status: Project Completed; 2017-2018.
Team: 4 members, PM; two information architects; data scientist

MY ROLE

I was the Information Architect  and Project Lead on my team, and:

• Created taxonomic and metadata schemas for UE4 documentation guides

• Create data governance guidelines for UE4 documentation guides

• Conduct Data Analysis of UE4 documentation guides

• Conduct User Interviews, Surveys, and Testing with UE4 documentation guides, past and redesign models

• Assist in management of overall project

SKILLS APPLIED

• Card Sorting

• Content Inventory

• Content Management

• Content Strategy

• Data Quality Testing

• Excel/Google Sheets

• Information Architecture

• Journey Mapping

INTRODUCTION

Creating a video game is an arduous process, especially in an ever-growing billion-dollar industry. Epic Games’ Unreal Engine 4 gives users the opportunity to develop their own games. However, their documentation often faces challenges of discoverability and usability due to non-standard tagging vocabulary. Our project solves this problem by creating a taxonomy, which allows users of different backgrounds and skill levels to easily search through documentation. It produces significant commercial implications by improving accessibility for new users, and serving as a foundation for future work, such as ontology development and Documentation as a Service, increasing developer participation and diversity.

CURRENT PROBLEMS

I would need to solve the following problems to ensure the redesign of the UE4 documentation functionality and navigability worked.

• Data analysis of UE4 Documentation

• Guidelines and use and accessibility

• Understand Users

• Competitor Analysis

• Organization and Structure Preference of UE4 Documentation

• User research and testing

Our team began this project by pinpointing the problem space we wanted to explore: Who are the Users?

Design question: How can we enable various user types to access, navigate, and use the UE4 documentation and gain something out of it?

IDENTIFY GOALS

Throughout the project I was responsible for the following tasks:

1. Analyzing UE4 Documentation for topics, categories, preferences, user types, proficiencies, game mechanics, guides, terminology and other attributes

2. Conduct User Interviews, Surveys, and Testing for current and redesign models of UE4 Documentation

3. Design taxonomy and ontology for UE4 Documentation

RESULTS

Our redesign incorporated several of the key areas outlined by our user research and data analysis. We wanted to approach our redesign by incorporating user types for skill levels and types of learners. We wanted to solve the problem of educating one type of user while still giving the freedom of mobility for our more advanced audience. With the incorporation of suggested and alternate tags (for subjects and topics) the intermediate and advanced users can locate information at their own pace, improving browsing and search capabilities in the UE4 documentation. We also provided recommendations for UX and UI improvements, making documentation functionality a positive when referring to usability experience and not making it harder.

ONGOING ACTIVITY

I am pleased to say we were the first group to work with EPIC games as outside consultants as part of a program with UW. We were able to leave the project, handing off suggestions to the next group who would come in and continue our efforts to improve the documentation redesign. Some of those recommendations were.

• Incorporate feedback from a new batch of users: I would reconstruct our usability tests, surveys, and interviews and add in a new dimension we missed to add a new area(s) to give attention to in the next iteration.

• Continue data analysis with python notebook: As we were finishing up the project I was working on a python notebook to look at clustering data of the UE4 documentation. The auto generated clusters could help provide another organizational way (semantically) we may want to structure the data and how it could help sort the data we have already classified. It could add to our data quality testing as well

• Add an ontology: We were only scratching the surface here. What I wanted to do and wish I could go back and try now is adding personalization based on the preferences we established from the user research. Using the user types and learning modalities we could create an ontology of subjects, topics, skills, and processes that work best for each type and create a faceted search for UE4 documentation utilizing those attributes.
   

KEY LEARNING POINTS

This project gave me the opportunity to work with a large dataset and a large user group that has several different needs, requests, and concerns when they want to access and use information from the documentation. I needed to use my whole bag of tricks on this project to understand how to manage a project with a concrete deadline, perform several user research tasks, engage with several stakeholders, and provide recommendations based on my data analysis and classification of the data. This experience helped me understand the values of time, communication, and other’s success. I needed to make sure everyone's points of views were heard, and ultimately guide us on the right path to success.

I was able to improve my skills in taxonomy and ontology development, database management, content inventories, competitor analysis, navigation, guidelines, and user flows. This project has taught me how to approach data quality testing when taking into account user research and testing data. I am excited to use the skills from this project to help shape other user experiences in other contexts.