IA Summit 12: Filtered and Refined
After the Better Cross Channel Experiences talk the next 20 minute talk for that session was Filtered & Refined: Interfaces for Distilling Data by Erin Jo Richey. This talk was all about the way that information is organized and presented to people with a focus on multifaceted search queries.
Dynamic multifaceted search queries provide us with specific answers to dynamic questions that aren’t well defined. They augment our cognition.
Types of Data
There are 4 types of data that a person might need when trying to identify or find information:
- Your are in the category, or you aren’t.
- You can’t do math on categorical data.
- Examples: shirt color, hair color.
- Items that have a natural order or hierarchy.
- You can’t do math on ordinal data.
- Examples: shirt sizes (S, M, L, XL), days of the week (S, M, T, W, T, F, S)
- You can perform numeric calculations on quantitative data
- Examples: prices, dimensions.
- Forms of numeric data with spatial relationships
- Examples: maps, coordinates.
This data can be represented many different ways. For example when representing what and where, use categories and spatial data like positions in space, shapes, and groups. When representing how much, use positions on a scale, saturation, area or size, and volume.
Interface controls can also be used to represent data, for example:
- Choose 1 = Radio Buttons
- Choose many = Check Boxes
- Choose Range = Slider
- Create Own = Text Box
- Remove Points = Individual Deletes
- Sort = Drop Downs
Creating a Whole
What brings all the pieces and parts together to create a good experience is an interface that allows a broad overview, with the ability to zoom and filter and then see details on demand.
Adding things like history, extraction, and marking can then take the interface to the next level.
- History – Giving people the ability to make choices and refine them without starting over.
- Extract – Having the ability to save results, email, export, share and print.
- Mark – Allowing the user to highlight, denote favorites, make notes, and flag items of interest.
These interfaces should help people make decisions, by giving meaning to the data. This meaning will help them see the patterns they are looking for.