Mathesar Media Asset Browser

Research Phase

Design Brief

Background

• Problem: The current Mathesar GUI only supports displaying textual data in the database tables. Media assets like images, videos, and audios stored as URLs in the database tables are shown as plain text (unclickable links) on the interface. Users whose work is media-driven and often involves browsing and editing media assets stored on the servers or cloud have this need of easily previewing and playing their media “on the go.”
• Solution: Create a design for a media-focused view for Mathesar. This new view aims to extend Mathesar by offering a convenient way to browse and manage media assets that are already stored in the user’s database tables.

​

Target User

• Mathesar: non-technical or even technical users (how to handle data where they can scale them and understand data modeling concepts)
• Media Browser: non-technical users with media asset management needs with their databases.

​

Context of Use

• User personas and use cases are assumptions based on clients’ vision and user research (see personas and storyboards section).
• Interactions and experiences will be refined after the study of users and use cases.
• URLs stored in the table will be displayed as their corresponding media types (e.g. URLs to images will be shown as thumbnails/pictures)
• Cache and loading experiences need to be considered.
• Be able to switch viewing contexts depending on levels of uses (e.g. pop-ups for light users vs. gallery views for heavy media users).

​

User Goals

• Make the technology more accessible with visual references.
• Make database management less daunting for non-technical users.
• Exploration of media: tables are not the best way to view all data types; could involve a media plug-in in the future.

Secondary Research

• Researchers are customizing tools like Airtable to share with non-technical users.
• “A copy of my database, with live data removed or replaced, can be accessed at https://bit.ly/RelicView (case sensitive). This provides read-only access to the full database. To view the database with full editing permissions, readers may create a free Airtable account, visit the above URL, and click “Copy base” in the top-right corner of the screen (“base” is Airtable’s shorthand for “database”). Using their own copies of the database, readers may edit and experiment further or simply view in detail the features described below” (Robert Heaton, 2020).

​

• 10 things non-technical users don’t understand about your software
• The difference between web and native applications
• Data storage
• The jargon you use
• Concurrency
• What changes can be reversed
• The need for backups

Secondary Research

• Hierarchical organization (especially when different attributes in one hierarchy) is problematic.
• Researchers found heavy collection users preferred multi-faceted metadata grouping of images to visually similar grouping (Yee et al., 2003).

Secondary Research & Competitive Analysis

• Media databases - online media inventory
• Unsplash: Beautiful Free Images & Pictures: online image inventory
• Images displayed in grid layout
• Associated metadata shown on hovering
• Click into the image to reveal more details
• Support searching and filtering
• Category-based tags (if a column in the database is an enum, can be used as tags displayed on the navigation bar)
• User-created tags – can be chaotic

Secondary Research & Competitive Analysis

• Media databases - online media inventory
• Vimeo | The world's only all-in-one video solution: online video library/storage
• Upload videos by dragging and dropping
• Titles displayed under video thumbnails
• Search based on video content
• Supports organizing and managing videos

Secondary Research & Competitive Analysis

• Media databases - online media inventory
• Storyblocks: online media inventory (audio, video, image)
• Grid layout
• Autoplay videos on hovering
• Click into the images/videos for more details
• Support searching and category-based tags

Secondary Research & Competitive Analysis

• Visual database tools - database extensions for managing and visualizing data
• Microsoft | Visual Database Tools: SQL Server Management Studio
• Dedicated solution for Microsoft SQL Server and Azure SQL Database
• GUI on top of the underlying databases similar to Excel
• Support displaying textual data only but with different data types (number, date, currency, etc.)
• Make database queries easy (creating table definitions, inserting/updating/deleting data and columns)
• Good reference for building on top of the existing table view

Secondary Research & Competitive Analysis

• Visual database tools - database extensions for managing and visualizing data
• DbSchema: The Best Free Database Designer & GUI Tool: GUI on top of databases
• Integration with multiple databases such as SQLite, MongoDB, Snowflake, etc.
• Table view with interaction achieved through mouse clicks
• Supports generating documentation for database table design
• No support for digital media

Secondary Research & Competitive Analysis

• Airtable
• Cloud collaboration service
• A spreadsheet-database hybrid, with the features of a database but applied to a spreadsheet. The fields in an Airtable table are similar to cells in a spreadsheet, but have types such as 'checkbox', 'phone number', and 'drop-down list', and can reference file attachments like images.

Secondary Research & Competitive Analysis

• Notion
• An application that provides components such as notes, databases, kanban boards, wikis, calendars and reminders. Users can connect these components to create their own systems for knowledge management, note taking, data management, project management, among others.

Secondary Research & Competitive Analysis

• Other products
• JotForm Tables
• Google launches a work-tracking tool and Airtable rival, Tables
• Microsoft launches Lists, a new Airtable-like app for Microsoft 365
• Asana: designed to help teams organize, track, and manage their work.
• ThoughtSpot: Search database with natural language queries
• Which Is The Best PostgreSQL GUI? 2021 Comparison (they all look so complex...)

Key Findings from Interviews

• Notion & Airtable users
• Enjoy the flexibility of using different levels of views (table, gallery, dashboard, etc.)
• Almost never solely rely on table views
• Add images in Airtable and change to card view for easier browsing
• Preview is always helpful, but can make the table look cluttered
• Mostly use these program for non-media type data – media is mostly a reference in differentiating data more easily

​

• Don’t use Airtable to house data often; typically uses the kanban view for task management at work
• When she does use Airtable to house data, what typically happens is she uses Google Sheets first and uploads the csv to Airtable
• For work, in the instances where she doesn’t use kanban view - the entry fields for the table are full sentences; not quantitative data

Key Findings from Interviews

• Airtable Designer:
• Airtable users do not consider Airtable a relational database; they recognize it as relevant because of what the tool does; perhaps it is a content calendar or project tracker.
• Users who set up a database scheme and layouts are designers in essence (not officially, and may not know it) because they are setting up the database to design a tool which solves their problem.
• The grid is the lowest common denominator because of its flexibility, but doesn’t suit all workflows.
• Views and layouts allow people to design their data around specific tasks/workflows in a more accessible way than coding a custom UI.

​

Key Findings from Interviews

• Content marketer
• Existing tools: Pinterest / Dribbble / Behance for visual design inspiration, file system / Google Drive for marketing campaign examples
• Share notes about content strategies with colleagues
• Need tools for campaign performance review and content analytics
• More than one media assets per record / row
• Pinterest / Dribbble / Behance users
• Infinite scroll interface helps create a smooth browsing experience
• Only support Internet resources (cannot just copy and paste screenshots)
• Lack of mixed media support e.g. mixing videos and images into a single mood board

Key Findings from Social Media Observations

• Non-technical users do need tools to manage or even build databases
• No great tools out there
• Many questions on social media are unanswered
• Couldn’t find results with keywords “non-technical ‘media’ database” on stackoverflow

Summary of Key Findings

• Use cases are important.
• Depending on the levels of use, media browsing views can be add-ons, plug-ins, or even the main views (rather than table views).
• The need of storing media data is not equal to the need of interacting with media data. The user experiences would be much different as well.
• View templates are helpful.
• Displaying textual data and media assets at the same time can be chaotic. View templates are helpful for non-technical users to start with their data.
• User experience needs to be carefully studied, especially under heavy use cases.
• Media data are usually stored elsewhere. Asset loading and user interactions while loading should be considered. (e.g., if they need to wait, what can they do in the meantime?)

Define

How might we improve the media asset browsing experience in the database for different levels of use, relative to the common table views?

• Media types: Images, videos
• User personas and use cases are assumptions based on clients’ vision and user research.
• Inventory (e.g., album or book covers); low level of engagement with media assets.
• Arts collection management; medium level of engagement with media assets.
• Journalists (e.g., photographs, footage); medium to high level of engagement with media assets.
• Non-technical roles at data collection companies (e.g., UX researcher); medium to high level of engagement with media assets.
• Researchers (e.g., computer vision researchers); high level of engagement with media assets.

Non-technical roles at data collection companies

ML Researchers

“I discover the secrets behind images through my machine learning models.”

​

About

James is a Ph.D. student in the EECS department at the University of Washington. His research focuses on computer vision and video engineering. He is going to enter the third year of his Ph.D. and aims for publishing his fourth research paper soon.

Goal

• Speed up the process of data cleaning for large image datasets.
• Organize and manage images and their corresponding labels with ease.
• Compare and categorize model predictions with ground truth side by side in a large scale.

Motivation

• It is common for large datasets to have missing/incorrect labels.
• Able to evaluate and understand model prediction errors makes it easier to correct and improve the model.

Challenges

• Data cleaning can only be done manually, which is time-consuming.
• It is relatively easy to save model predictions as images and store pairwise labels in a csv file but shockingly tedious to visualize and navigate images with labels in a large scale.

Investigative Journalist

“My research uncovers the truth about human rights violations, and is highly sensitive to the victims.”

​

About

Myra is an investigative journalist at The Quincy Institute for Responsible Statecraft. Her research focus on human rights violations and war crimes throughout the Middle East and North Africa.

​

Goal

• To analyze video content in a time efficient way while taking organized and structured notes
• Identify the timestamps and screen caps in sensitive videos with artifacts pertinent to the project goal

​

Motivation

• Safely record and analyze without risk
• Spreadsheets do not fully allow for customization

​

Challenges

• The information dealt with is highly sensitive and must be stored in a database that is encrypted
• Encrypted databases are complex and I am unaware of how to use them; someone else handles this

Museum Curator

“Curating an exhibition can take 2 to 3 years. The process involves countless explorations to articulate the themes and connections among artifacts.”

​

About

Alicia is a curator assistant at the Berkeley Art Museum and Pacific Film Archive. She helps curators at the museum manage their art collections and communicate with the artists.

Goal

• To manage logistics, such as the budget and the loan arrangement with artists
• To explore the relations and themes of the artifacts

Motivation

• To design the best layout and experience for the exhibition
• To present the intent of the artists accurately

Challenges

• No single place for exploring collections and managing loan
• Difficult to look at the overview of multiple artifact together