Abstract:Purpose - The algorithm of disjoint literature-based knowledge discovery provides a convenient, efficient and effective auxiliary method for scientific research. Based on an analysis of Swanson's A-B-C model of disjoint literature-based knowledge discovery and Gordon's intermediate literature theory, this paper seeks to propose a more comprehensive compound correlation model for disjoint literature-based knowledge discovery. Design/methodology/approach - A new algorithm of vector space model (VSM) based disjoint literature-based knowledge discovery is designed to implement the compound correlation model. Findings - The validity tests showed that this new model not only simulated both of Swanson's early and well-known discoveries of Raynaud's disease-fish oil and migraine-magnesium connections successfully, but also applied to knowledge discovery in the agricultural economics literature in the Chinese language. Research limitations/implications - Although the workload was reduced to the minimum under the compound correlation model compared with other algorithms and models, part of the work needed some manual intervention in the process of disjoint literature-based knowledge discovery with the VSM-based compound correlation model. Practical implications - The algorithm was capable of knowledge discovery with a large-scale dataset and had an advantage in identifying a series of hidden connections among a set of literatures. Therefore, application of the model might be extended to more fields. Originality/value - Traditional two-step knowledge discovery procedures were integrated into the model, which contained open and closed disjoint literature-based knowledge discovery.

Generalized Theme Dictionary Models for Association Pattern Discovery

Association Pattern Discovery Via Theme Dictionary Models

A Classification-Oriented Dictionary Learning Model: Explicitly Learning the Particularity and Commonality Across Categories.

A dictionary learning approach for classification: separating the particularity and the commonality

Class-Aware Analysis Dictionary Learning for Pattern Classification.

Latent Theme Dictionary Model for Finding Co-occurrent Patterns in Process Data

Discovering treatment pattern in traditional Chinese medicine clinical cases using topic model and domain knowledge

Modeling the Evolution of Associated Data

A Compound Correlation Model for Disjoint Literature-Based Knowledge Discovery (vol 64, pg 423, 2012)

Topic Models Incorporating Statistical Word Senses

Fast, Flexible Models for Discovering Topic Correlation across Weakly-Related Collections

Statistical Word Sense Aware Topic Models

Probabilistic Topic Modeling for Comparative Analysis of Document Collections

Motif-based Rule Discovery for Predicting Real-valued Time Series

Generalized Relational Topic Models with Data Augmentation.

A Model for Combinatorial Dictionary Learning and Inference

Modeling Dynamic Topics in Chain-Free Fashion by Evolution-Tracking Contrastive Learning and Unassociated Word Exclusion

On-line supervised theme-modeling and evolution-analyzing method

Association Rule Mining Based On The Multiple-Dimensional Item Attributes

A Tale of Three Probabilistic Families: Discriminative, Descriptive and Generative Models

Generalizing Case Frames Using a Thesaurus and the MDL Principle