Progressive Learning of Topic Modeling Parameters: A Visual Analytics Framework
MetadataShow full item record
Topic modeling algorithms are widely used to analyze the thematic composition of text corpora but remain difficult to interpret and adjust. Addressing these limitations, we present a modular visual analytics framework, tackling the understandability and adaptability of topic models through a user-driven reinforcement learning process which does not require a deep understanding of the underlying topic modeling algorithms. Given a document corpus, our approach initializes two algorithm configurations based on a parameter space analysis that enhances document separability. We abstract the model complexity in an interactive visual workspace for exploring the automatic matching results of two models, investigating topic summaries, analyzing parameter distributions, and reviewing documents. The main contribution of our work is an iterative decision-making technique in which users provide a document-based relevance feedback that allows the framework to converge to a user-endorsed topic distribution. We also report feedback from a two-stage study which shows that our technique results in topic model quality improvements on two independent measures.
- Faculty Publications 
Showing items related by title, author, creator and subject.
A flexible, longitudinal and surrogate consent model: Consent of Infants for Neonatal Secondary-use research (CoINS) Model Choi, Yvonne (2020-04-01)Documenting healthcare, along with technology enabling capture of streaming patient telemetry, can deliver large datasets offering opportunities to discover new insights primarily identified through retrospective secondary ...
Developing a thermodynamic model for the U-Pd-Rh-Ru quaternary system for use in the modelling of nuclear fuel Wang, Lian Cheng (2018-12-01)Ruthenium, rhodium, and palladium are fission products in nuclear fuels. These elements and their compounds change the properties of fuel pellets. Phase diagrams involving uranium have been constructed experimentally to ...
Alshehri, Abdullah (2012-07-01)The purpose of this thesis is to analyze a non-intrusive connectivity visualization method for OLSR-based MANET topology in different mobility models. The visualization relies on the local topology databases (neighborhood ...