Weimao Ke

Weimao Ke

Auto-generated profile, based on publicly available information

Weimao Ke

Indiana University

Research Focus & Metrics

Academic Career Length

20Years (2007-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 July 2007 in 2007 11th International Conference Information Visualization (IV '07)
Verified
Conference Paper
Closed access
BHBruce W. Herr
Bruce W. Herr
WKWeimao Ke
Weimao Ke
EHElisha Hardy
Elisha Hardy
KBKaty Borner
Katy Borner
This paper presents the results of an analysis and visualization of 428,440 movies from the Internet Movie Database (IMDb) provided for the Graph Drawing 2005 contest. Simple statistics are presented as well as a tapestry of all movies with an overlay of the giant component of the co-actor network. Academy award winners are highlighted. Major insights are discussed.

16 July 2021
Verified
Thesis
Closed access
XGXuemei Gong
Xuemei Gong
WKWeimao Ke
Weimao Ke
Classic information retrieval (IR) systems rely on ranking algorithms to serve users with ordered lists of documents according to search queries. Sometimes, however, users do not have very specific information needs or cannot accurately articulate their information needs in queries. Cluster-based IR systems, such as those based on the Scatter/Gather paradigm, have been used to help users clarify their information needs and promote learning via interactive document clustering and summarization. These systems have the potential to facilitate user browsing large document collections and exploring topics. However, their effectiveness is often constrained by poor clustering quality, ambiguous cluster labels, and the inefficiency to process large-scale data sets. In interactive clustering, term distributions vary in different clusters or subsets of a collection. Classic TF*IDF (term frequency * inverse document frequency) term weighting, especially IDF that counts document frequency in the overall (global) data, does not take into account the shifted term distributions in a (local) subset and is often incapable of identifying most informative terms within that subset. To improve clustering quality with meaningful labels, we propose two novel term weighting schemes, namely TF*ICDF and DF*LIG. TF*ICDF, or Term Frequency * Inverse within-Cluster Document Frequency, integrates the local subset information into term weighting. It outperforms TF*IDF in several aspects for clustering and labeling with various configurations. In addition, we propose Least Information Gain (LIG) based on the least information theory, which, similar to Information Gain (IG) based on KL divergence, measures the amount of information required for a probability distribution change. Based on LIG, we develop the DF*LIG method for cluster labeling. With DF*LIG, terms that carry more information in revealing the contents of clusters are chosen as labels, resulting in better performance in terms of coverage, overlap and precision in comparison to DF*IG. By integrating TF*ICDF for term weighting and clustering, DF*LIG produces more representative, distinctive and accurate labels than when it is combined with TF*IDF. In order to improve clustering efficiency and support data-intensive processing, we develop distributed versions of TF*ICDF and DF*LIG algorithms as well as a parallel clustering algorithm named Pruned Affinity Propagation (PAP) in the Spark framework. The proposed algorithms efficiently process large-scale data sets by taking advantage of computational capabilities of individual processors and nodes. Distributed TF*ICDF and DF*LIG methods scale very well - their efficiency improves significantly with an increased number of processors. Compared with the original affinity propagation algorithm, PAP achieves much higher efficiency while maintaining strong effectiveness. Results also show that the execution time of PAP is greatly reduced by increasing the number of processors and remains competitive with large numbers of documents, indicating its scalability. With the support of these effective and scalable methods for text clustering and cluster labeling, a cluster-based IR system can be greatly improved in its ability to dynamically identify key features, to produce meaningful clusters, and to generate representative terms as labels. With the ability to accommodate large-scale data sets, such a system can help users discover important patterns in the data and help them learn and explore in a dynamic, complex information space.

6 July 2021
Verified
Thesis
Closed access
BSBo Song
Bo Song
XHXiaohua Hu
Xiaohua Hu
WKWeimao Ke
Weimao Ke
The amount of available data has experienced significant growth as the result of technology advances in this era of Big Data. The biomedical domain, in particular, is one exemplar field where the number and scale of data sources have increased exponentially in the last decade. They are expected to keep growing even more rapidly to reach the level of Zetta bytes per year in the following year very soon. While more data obtained from advanced biotechnologies, such as high-throughput sequencing that encodes valuable information, are becoming overwhelmed, to discover knowledge from which for biology and medical research is still facing challenging problems with existing approaches. We study in this dissertation how to effectively and efficiently utilize these large-scare data from different numbers and types of sources for biomedical knowledge discovery. Raw data from biological organism such as microbiome usually have intrinsic high dimensionality of the feature space, which inevitably and exponentially raises the computational complexity of existing algorithms. We proposed a new approach using spectral interpolation technique to represent the high-dimensional data in low dimension space that not only greatly improves the efficiency of computing large-scale data but also preserves as much information as possible from original data. The resulting preferable outcomes for clustering and visualization tasks better facilitate the knowledge revealing of patterns and insights for microbial communities. We further studied how to enhance knowledge discovery while more than one data sources are available. Large-scale relational data such as protein-protein interactions (PPI) can be constructed in the form of network to invoke a system-wide perspective than traditional mechanistic approaches to interpret complex biological processes and functionalities. While bio-experiments are exhausted and costly, with two or more networks from different data sources we can apply computational comparative analysis such as Network Alignment to bridge the knowledge between well-studied species and under-examined species. We proposed new methods to globally align multiple large-scale biological networks from different species at the same time. We utilize both topological features and biological features of PPI networks and search heuristically for the best results. Representation learning for network is also integrated into our proposed framework to provide a new way to quantify the structural features of a node with its surrounding topology for the node embedding. The real data experiments showed promising results in finding homologous proteins as well as conserved protein complexes in poor-studied species for knowledge transferring from well-studied species. Besides utilizing homogeneous data from one and more data sources of one type, we keep exploring the possibility of harnessing sources of different types to take advantage of their underlying relational knowledge across heterogeneous data and capture the complex biomedical associations. The heterogeneous disease information networks we formulated in one research include types of sources from disease, pathway, and chemicals. They are filtered and calculated using Dynamic Time Warping (DTW) algorithm and meta path method for topological and semantics scores which lead to effective measurement of the similarity of diseases. In another study, we proposed a novel framework with Graph Convolutional Network to identify and predict disease-RNAs associations to better support the discovery of relational knowledge at the molecular level for medical applications such as disease diagnosis, therapy, and monitoring.

30 June 2025
Verified
Thesis
Closed access
SPSonia M. Pascua
Sonia M. Pascua
WKWeimao Ke
Weimao Ke
This dissertation presents a novel information-theoretic framework, ROADMAP (Representation, Organization, and Analysis for Data Modeling and Annotative Predictions), for advancing information representation, knowledge organization, and predictive analytics in machine learning. Motivated by emerging limitations in classical entropy-based approaches, this research explores how new formulations in information theory can lead to improvements in accuracy, robustness, interpretability, and annotation efficiency in AI systems. Central to this investigation is DLITE (Discounted Least Information-Theoretic Entropy), a new entropy metric grounded in Least Information Theory (LIT), a mathematically rigorous formulation designed to quantify bounded entropy change, normalize scale effects, and satisfy essential information-metric properties - providing both theoretical clarity and practical utility serving as an alternative entropy-based quantification model. DLITE was formulated by Ke (2020, 2022a) to address key theoretical and practical gaps in traditional information theory models such as Shannon Entropy and KL Divergence. Unlike popular information-theoretic measures, which lack metric distance properties, DLITE introduces a bounded and symmetric measure of the entropy difference, discounting redundant or scale-sensitive information using a formal entropy discount. By satisfying properties such as nonnegativity, symmetry, and identity of indiscernible, DLITE provides a scale-invariant, semantically meaningful framework that is theoretically rigorous and computationally robust. Additionally, central to this study is the conviction that a structured framework is essential to effectively address complex problems in AI. The ROADMAP framework, a three-tier model, ensures that each stage of information processing is logically aligned, systematically implemented, and analytically evaluated. This framework provides a scaffolded strategy for integrating theory with methodology, enabling deeper insights and reproducible experimentation. Moreover, this study operationalizes DLITE within a tiered ROADMAP framework. In Tier 1, DLITE informs feature weighting via TF-iDL, an entropy-discounted alternative to TF-IDF that reflects more meaningful semantic salience. In Tier 2, DLITE Impurity Measure (DIM) is integrated into decision tree models, resulting in more stable and informative node splits for hierarchical knowledge classification. In Tier 3, DLITE Loss is deployed as a loss function in transformer-based deep learning models, offering strong recall, fast convergence, and resilience to class imbalance, as evidenced in comparative experiments against Cross-Entropy and KL Divergence on CoNLL-2003, Basic NER, and the Broad Twitter Corpus. Beyond empirical performance, DLITE supports selective annotation strategies and enhances explainability through interpretable attention and entropy landscapes. It offers a new way of thinking about how learning systems evaluate and optimize information, especially under conditions of noise, ambiguity, and limited data. This dissertation demonstrates that DLITE is not just a derivative of existing information theory but a substantive contribution to it: one that redefines how entropy is modeled for learning, ranking, and reasoning in machine intelligence. But this research is also personal. It is not only a pursuit of technical advancement but a tribute to intellectual lineage. Like the storied advisor-advisee relationships that shaped entire disciplines--Frege and Wittgenstein, Thomson and Rutherford, Hilbert and Von Neumann--this dissertation emerges from a legacy of mentorship. The core theories at its heart, LIT and DLITE, were conceptualized by my advisor, Dr. Weimao Ke. My work, then, is not only a scholarly inquiry, but a continuation--a living expression of the ideas entrusted to me. This dissertation exemplifies how new knowledge can emerge from lineage, shaped by the values of academic stewardship and inspired by the responsibility to carry forward a vision. In the tradition of advisor-advisee collaborations that have moved science and scholarship forward, this work aims not just to validate a theory, but to extend its reach. DLITE is more than a mathematical model; it is an invitation to rethink how we quantify, interpret, and apply information. In that spirit, this dissertation contributes to the evolving landscape of explainable, human-aligned AI, and to the scholarly lineage from which it was born.

16 July 2021
Verified
Thesis
Closed access
YZYizhou Zang
Yizhou Zang
XHXiaohua Hu
Xiaohua Hu
WKWeimao Ke
Weimao Ke
With the ever-growing volume of online information, recommender systems have become an effective strategy to reduce information overload. Collaborative Filtering (CF) is a widely used technique by modern recommender systems. CF uses the observed preferences/behaviors of a group of users to make predictions on the un-known preferences/behaviors of other users. Direct rely on user behaviors allows CF to uncover complex and unexpected patterns without understanding item content. While generally more accurate than other recommendation approaches, such as Content-based recommendation, Collaborative Filtering suffers from data sparsity and cold start problems due to its lack of capability to deal with limited user-item interaction data.