The first version of MediSyn synthesizes two drug-target datasets. It uses a matrix-based layout to visually link drugs, targets (i.e., mutations), and tumor types. Data uncertainties are salient in MediSyn as (i) missing data are exposed through the matrix view; (ii) inconsistencies between datasets are shown via overlaid layers; and (iii) data credibility is conveyed through links to data provenance.
The improved version of MediSyn synthesizes five drug-target datasets. We also redesigned the interaction of MediSyn based on the notion of entities (that is, taking drugs and targets as entities) for insight generation. MediSyn supports five types of interactions: 1) selecting entities of interest, 2) connecting relevant entities, 3) elaborating by retrieving the details of entity relations, 4) exploring other entities, and 5) entity-based insight sharing.
Tuple retrieval visualizes search results under various combinations of search terms. It also suggests relevant terms and combinations of terms to help navigate through the search space.
IntersectionExplorer applied the design of UpSet to inter-relate publications generated by recommendation agents and other users (such as bookmarks) with the idea that the item suggested by multiple sources can increase the likelihood of user adoption of it. Source code
There are a series of extensions I made for the dynamic and flexible interactive photoshow (D-Flip) as my master's research project. The extersion includes adapting the system to various types of content, such as paintings and flashcards, and to various interaction modalities, such as multi-touch public displays and whole body interactions. More videos
Chen He, Luana Micallef, Liye He, Gopal Peddinti, Tero Aittokallio, and Giulio Jacucci, Characterizing the quality of insight by interactions: A case study, IEEE Transactions on Visualization and Computer Graphics (TVCG). [PDF]
Chen He, Luana Micallef, Zia-ur-Rehman Tanol, Samuel Kaski, Tero Aittokallio, and Giulio Jacucci, MediSyn: Uncertainty-aware visualization of multiple biomedical datasets to support drug treatment selection, BMC bioinformatics, 18, No. 10, 393, 2017. [Open access]
Chen He, Denis Parra, and Katrien Verbert, Interactive recommender systems: A survey of the state of the art and future research challenges and opportunities, Expert Systems with Applications, Vol. 56, 1 September 2016, pp. 9–27, ISSN 0957-4174.
Katrien Verbert, Karsten Seipp, Chen He, Denis Parra, Chirayu Wongchokprasitti, and Peter Brusilovsky, Scalable exploration of relevance prospects to support decision making, Proceedings of the Joint Workshop on Interfaces and Human Decision Making for Recommender Systems co-located with ACM Conference on Recommender Systems (RecSys 2016), Vol. 1679, CEUR Workshop Proceedings, 2016.
I teach the Interaction Data Visualization course at University of Helsinki as part of the Master's Program in Data science and Computer Science. Check out the lecture slides below.
As an introduction to data visualization, this lecture discusses what is visualization, why do we need visualization, and what are the purposes of visualization. It also introduces several common techniques of visualizing multi-dimensional, graph, and set-typed data. Download
Topics include optical illusions, visual processing model, preattentive processing, gestalt laws, etc. Download
This lecture talks about the nested model for visualization design proposed by Tamara Munzner, the LATCH principle, and Tufte design principles. Download
This lecture talks about techniques of visualizing one, two, and multi-dimensional data, as well as the use of composite views for multi-dimensional data representation. Download
This lecture talks about techniques to visualize graphs (e.g., node-link diagrams and adjacency matrix) and trees (e.g., space-filling techniques). Download
This lecture is about techniques to visualize set-typed data, such as Venn and Euler diagrams. Download