TASK 1: Static Overview of 10 years of Infovis
-
Process:
We used technology developed internally at Microsoft Research to cluster the papers in the InfoVis proceedings. In order to cluster, we used the titles, references and keywords (if available) in the clustering process, weighting the titles more heavily to get a better clustering result. A standard list of stop words, months of the year, journal and proceeding titles and version and page numbers were removed from influencing the cluster results. -
Image 1 :
-
Insight:
Five clusters emerged from using the clustering software:- General (58 papers)
- Dynamic Queries (28)
- Graph Visualization (19)
- Focus + Context techniques (16)
- Tree Visualization (31)
The General category is the most popular every year except for 2 of them: Tree visualizations were very popular in the year 2000 and Graph Visualization papers were most common in 2001. Also, the topic of Graph Visualization has grown in popularity quite recently, while the topic of Dynamic Queries has enjoyed a steady increase over the last 8 years.
CHI publications have been the main source of inspiration for the InfoVis community to date. In the top 10 papers most frequently cited, there is only one InfoVis publication. -
Caption for exhibit:
Eight Years of InfoVis Conferences
Note: Checking more than one checkbox will highlight papers that match
either criteria. PaperLens has also been extended to visualize 23 years (1982-2004) of the ACM
SIGCHI conference proceedings.
TASK 2: Characterize the research areas and their evolution
Same as Task 1.
Described above.
The popularity of topic view can be used
for both input and output for Task 4, "Which papers/authors are
most often referenced?" The user can quickly discover not only
the most influential researchers/papers in a particular topic area but also
which topic areas were most influenced by the selected paper/author.
TASK 3: The people in InfoVis
Task 3.1: Where does a particular author/researcher fit within the
research areas defined in task 2?
authors
list , they are shown in the selected
authors area and added to the From combo-box in the
degrees of separation links view. Once authors are added
to the selected authors area, all of the
papers by those authors are also shown in the paper
list and are highlighted in the popularity
of topic view, matched to the author by color coding. When two
or more selected authors have contributed to a paper, we use black and show the
number of authors from the selected authors on the paper.
When the user selects authors from the
Selecting an author from the year by year top 10 cited
authors view shows the papers that author has published in
InfoVis in the popularity of topic view.
And, the papers that have referenced that author are highlighted in orange in
the same view.
When an author is selected from the From combo-box, we display all
of the related colleagues to the selected author in the To combo-box
with the corresponding degrees of separation. When an author is selected
form the To combo-box, we display the shortest path between two authors
in our dataset through their degrees of separation links
.
George Robertson has published only one paper, "Constellation: A
Visualization Tool for Linguistic Queries from MindNet" at InfoVis with
Munzner and Guimbretiere. So, he fits in the Dynamic Queries
topic area.
George is the 4th most frequently cited author by the InfoVis community.
He has 3 related colleagues (but not co-authors) through these 2
co-authors. His papers have been highly referenced by papers in the
Focus+Context and Tree Visualization topic areas.
Huub van de Wetering has published three papers at InfoVis. All of
them are in the Tree Visualization topic area. We could confirm
that "Visualization of State Transition Graphs" paper is also related to
tree visualization by reading the abstract at the ACM Portal.Task 3.2: What, if any, are the relationships between two or more
or all researchers?
Same as Task 3.1.
George Robertson and Stu Card have never published together at
InfoVis. However, two papers co-authored by them are in the top 10 most
frequently cited papers. Both of them were published in the CHI
proceedings.
Staffan and Johan wrote two papers together in the Focus+Context
topic. Staffan was the first author for both papers. They have
published with Lars Erik Holmquist together in one paper.
Carriere and Kazman wrote two papers together. They have not published any
other papers.
Hix and Crabb are connected indirectly to each other by 5 papers and 4
co-authors.
The relationships among all researchers: Not Addressed
Task 4: Which papers/authors are most often referenced?
popularity
of topic view, the year by year top 10
citations area is filtered to only show the frequent citations
for that topic area. Selecting an author from the year by year top 10
cited authors view (Figure 1g) shows not only any papers selected authors have
published in InfoVis, but also those papers that have referenced them via
orange highlighting in the popularity of topic
area. Hence, the user can quickly discover not only the most influential
researchers/papers in a particular topic area but also which areas were most
influenced by the selected author.
The top 10 overall most frequently cited references, which may or may not be
from the InfoVis proceedings, are viewable in the Overall list.
One can also see the detailed information for each of the top 10 papers
and how it was referenced over time by hovering on an individual paper.
When the user selects a topic from the
A double-click on an author in the year by year
top 10 cited authors view opens a number
of citations
view, which shows the number of citations of the papers written by the selected
author.
CHI publications have been the main source of inspiration for the InfoVis
community to date. In the top 10 papers most frequently cited, there is
only one InfoVis publication.
The most frequently cited paper by InfoVis authors to date is the Cone Trees
paper by Robertson et al., which was published in the CHI proceedings.
The majority of authors in the top 10 most frequently cited authors has
published at InfoVis. In the top 10 papers most frequently
cited, there are two authors who have never published at InfoVis.
S. K. Card was the most frequently cited author by InfoVis community to date
and the Cone Trees paper was the seminal contribution.
George Furnas was mainly referenced by the papers in the Focus+Context and
Tree Visualization topic areas.
Peter Eades was the most referenced author in 2001 when graph
visualization papers were most common even though his overall rank is
#15. Furthermore, he has never been in the top 10 except for that year.
For the focus plus context topic area, the fisheye paper by George Furnas was
ranked #1. It has never been referenced by papers in the dynamic queries
topic area.
For the tree visualization area, the Cone Trees paper by Robertson et
al. was the most referenced paper. It has been mainly
referenced by papres in that topic. Furthermore, it was ranked #1
for every year.
For the general area, Tufte was the most frequently cited author.
For the dynamic queries topic area, Ben Shneiderman was ranked #1.
For the graph visualization area, Peter Eades, who first introduced the spring
model to graph drawing, was the most frequently cited author.
For the focus plus context topic area, George Furnas and S. K. Card were ranked
#1 together.
For the tree visualization area, S. K. Card was ranked #1.
While Eades and Furnas were referenced by papers in their own areas,
Shneiderman and Card were referenced by papers in a broader range of topics.
Task 5: How many papers conducted a user study?
popularity of topic view, we highlight papers according to the
current selection.
When the user selects/unselects the check-boxes in the
Overall, only a small number of papers conducted a user study. However,
there are more papers with user studies in later years.
Most of the papers in the graph visualization and focus plus context topic
areas did not conduct a user study.
COMMENTS
