Thread Closed 
Week 11: The Flavor Connection
11-03-2014, 11:53 PM
Post: #1
Week 11: The Flavor Connection
The visualization can be found here:

[Image: The-Flavor-Connection-750x699.jpg]

Scientist have linked common flavored compounds across the world's favorite ingredients.
-Each blue dot represents a food item
-Similar foods are grouped into 14 category columns
-Bigger the dot, more popular the food among recipe
-Lines connect food that have compound in common (thick lines means more compounds are shared)
-Vertical position of the dot represents the total number of food connected to it and that it shares common compounds with bunch of other foods. Foods at the bottom are unique
-On selecting a particular food, we get a graph on the right which gives us a number of times that particular food has appeared in the recipe database. And on the left, the "number" of foods that shares a common compounds with the selected one.

-The viz is SLOW
-On filtering/selecting a particular food, the viz works smoothly
-The use of gray lines to connect food in same category was unnecessary.
-The graph on the right, which portrays the count of the food item in a recipe database, is too light. It didn't really catch my eye initially. Use of brighter color for left/right graph would have helped.
-The bottom of the graph looks very cluttered
-Because of space constraints, only the most popular ingredient in a cluster of dots is labeled. I wish there was a zoom-in/zoom-out effect. This would have helped to look through the cluttered space properly

Overall, it was a good viz. The graph looks subtle but the use of brighter color would have helped even more.
Find all posts by this user
11-05-2014, 09:00 PM
Post: #2
RE: Week 11: The Flavor Connection
Good critique!
I totally agree that use of gray lines to connect food in same category was hard to perceive if there were more food items.

The visualization looks neat and complex over all but it was hard for me to get my relations between items from inter-crossing lines.
I felt that when there are overlapping blue circles, I was not able to click and select a specific food item to view related items.
I was not able to compare between the thickness of the red lines connecting the food item. I would say this size encoding could have been replaced with colors that represent a range of number of shared compounds like 0-50 in yellow,red etc. An interactive legend would be useful for this purpose as well.
Find all posts by this user
11-09-2014, 02:17 PM
Post: #3
RE: Week 11: The Flavor Connection
Each dot is a food where the size channel is used to encode popularity of that particular food item. Lines connect foods that share similar flavor compounds. Color is used categorically to determine if two shared compound foods are in the same category or not, and size of the line tells how many compounds are shared.

The graphic starts out with an explanation page which I think was very clear and a great idea. When i first looked at the graphic I was overwhelmed, but the explanation cleared things up greatly. I don't think that the use of line thickness was a very good choice to show the amount of flavors shared. I could only really make qualitative estimates of how many flavors there were in a connection. I never knew the number exactly. I think the filtering interaction was very good. It really helped me see what was going on. However, I think a search bar would be super helpful. I wanted to look at some specific ingredients, like blueberries but I had to manually comb the visualization. If I could search for a term this would go a lot faster. All in all a fun visualization that allowed for good exploration of the data.
Find all posts by this user
11-09-2014, 06:15 PM
Post: #4
RE: Week 11: The Flavor Connection
I always feel like if a visualization needs explaining there's probably a better way of going about doing it. That being said, it was interesting to click through the various levels and see what was connected. I found that as you worked your way down the screen it became harder and harder to follow as it got more and more cluttered. I agree with the users above that the grey lines where wholly redundant.
Find all posts by this user
11-10-2014, 12:37 AM
Post: #5
RE: Week 11: The Flavor Connection
I totally agree with Greg and I think this visualization really needs a big guidance so you can get anything of it. Although I am not sure how to encode so much details in a better way but for me the graph was too cluttered. Graph started to be slow initially but was better when there was a selection. I also agree on unnecessary use of gray lines. Generally I didn't like graph and it wasn't visually appealing to me mostly due to it cluttering.
Find all posts by this user
11-10-2014, 12:05 PM
Post: #6
RE: Week 11: The Flavor Connection
I liked your critique. I agree that the chart on the side was not as noticeable as I would have liked it as well. When I first saw all the different lines and dots, I was really confused. I didn't even see all the dots at first. It took some time to see that they were related in some manor. When I selected a node it seemed to make it much more readable. I would say that the interactivity was essential for this visualization. I also feel that the visualization also shows that there could have been more interactivity. I agree there should be a zoom feature, it would help to see the more clustered items.
Find all posts by this user
11-10-2014, 08:24 PM
Post: #7
RE: Week 11: The Flavor Connection
As soon as you click the link the help widget displays the rubric on "how to read this graphic" and what each bubble or feature in the graphic represent. This is a neat thing to have as the graphic looks quite complex. I agree that the Zoom in/out feature would have helped. The grey lines complicate the graphic.The click to select feature shows clustering of items.
The bottom of the graph looks cluttered due to space constaints.
Interactivity could have improved this graph greatly.
Find all posts by this user
11-10-2014, 09:42 PM
Post: #8
RE: Week 11: The Flavor Connection
Crossing lines makes it very difficult to interpret the relation between different food items. User has to read through the explanation page to understand the vis, this makes it less effective. The visualization should allow user to interpret the data or understand the story on its own. To de-clutter the graph the user should have a zoom-in interactivity such that it allows him to zoom into the region of interest without which the graph especially at the bottom is extremely cluttered.
Find all posts by this user
11-17-2014, 05:44 PM
Post: #9
RE: Week 11: The Flavor Connection
I think its a fun visualization. While there are way too many line crossings, I do feel the authors have attempted to simplify this - for instance, by including transparency, or graying out some of the less important lines. I agree with pretty much everyone here that a zoom in functionality would have been beneficial, but it is somewhat implicit in the "click" a row to select or deselect a level/cluster of the graph. A zoom in feature would have lost the "global picture" and connections, which are viewed easily by simply selecting the desired cluster.
Find all posts by this user
Thread Closed 

Forum Jump:

User(s) browsing this thread: 1 Guest(s)