"If we knew what it was we were doing, it would not be called research, would it?" Albert Einstein
Kodzo Wegba, PhD Student
I am a Ph.D. student in the Department of Computer Science at University of North Carolina at Charlotte. My advisor is Dr. Wlodek Zadrozny. My research interest includes Natural Language Processing, Machine Learning and Text Mining. I specifically focus on interactive text summarization with "user in the loop".
I am a member of the UNCC Text Analytic Lab.
​
2019 Conference of the ACL will be in Fl0rence
7.28.2019 - 8.02.2019
ICPR 2019
07.08.2019 - 07.10.2019
UPCOMING EVENTS
MY LATEST RESEARCH
Discovery of Rating Fraud with Real-Time Streaming Visual Analytics
Examples of co-mapped SVD diagrams. Each row shows different mappings of one dataset. Comparing the co-mapped SVD diagrams in columns (c) - (f) with the direct SVD projections in columns (a) and (b), the co-mapped SVD diagrams avoid the misinterpretation of mapping independent U and V spaces directly, reveal edges between users and items with less clutters, and preserve the grouping information on the circle layout. For example, the two outlier movies shown in (d) of the first row are hidden in (a) and (b) and the connections between groups of users and items are generally better revealed in (c)-(f) than the original projections in (a) and (b).
Interactive Movie Recommendation Through Latent Semantic Analysis and Storytelling
An example from our latent semantic model for interactive recommendation and abstraction of user preferences. Our approach identifies a 2D visualization domain, where the horizontal axis layouts recommendable movies on a latent dimension between two combined movie features that are selected based on the user’s watch history, and the vertical axis uses recommendation degrees to move highly recommendable movies to the top. This example demonstrates the preference of a user on drama/documentary/biography movies (green zone toward the right) over comedy/music genres (orange zone toward the left). The movies selected to recommend are enlarged as blue circles, recommendable movies are shown as purple nodes, watched and liked movies as green nodes, and disliked movies as orange nodes. Two example movie posters, one liked movie “Casino” and one disliked movie “Airheads”, are also provided to demonstrate the latent dimension. For illustration purpose, we also add the arrowed line at the bottom and several movie titles to confirm the movie distributions on the visualization domain.
(a) (b) (c) (d) (e) (f)