Multi-Modal Bayesian Embeddings for Learning Social Knowledge Graphs

Authors: Zhilin Yang, Jie Tang, William Cohen

IJCAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on three datasets show that the proposed method clearly outperforms state-of-the-art methods. We then deploy the method on AMiner, an online academic search system to connect with a network of 38,049,189 researchers with a knowledge base with 35,415,011 concepts. Our method significantly decreases the error rate of learning social knowledge graphs in an online A/B test with live users.
Researcher Affiliation Academia Zhilin Yang Jie Tang William Cohen Tsinghua University Carnegie Mellon University jietang@tsinghua.edu.cn {zhiliny,wcohen}@cs.cmu.edu
Pseudocode Yes Algorithm 1: Model Inference
Open Source Code No The paper does not provide a specific link or explicit statement indicating that the source code for the described methodology is publicly available.
Open Datasets Yes We deploy our algorithm and run the experiments on AMiner1, an online academic search system [Tang et al., 2008]. ... We use the publicly available English Wikipedia as the knowledge base Gk. ... We use the full-text Wikipedia corpus2 as the text information C to learn the knowledge concept embeddings. Footnote 1: https://aminer.org/ Footnote 2: https://dumps.wikimedia.org/enwiki/latest/
Dataset Splits Yes Instead, we consider two strategies offline evaluation on three data mining tasks and an online A/B test with live users. ... For each researcher, we first compute the top 10 research interests provided by the two algorithms. Then we randomly select 3 research interests from each algorithm, and merge the selected research interests in a random order. When a user visits the profile page of a researcher, a questionnaire is displayed on top of the profile. ... We collect 110 questionnaires in total, and use them as ground truth to evaluate the algorithms.
Hardware Specification Yes The experiments were run on Intel(R) Xeon(R) CPU E5-4650 0 @ 2.70GHz with 64 threads.
Software Dependencies No The paper does not specify software dependencies with version numbers (e.g., Python version, specific library versions like TensorFlow, PyTorch, or scikit-learn versions).
Experiment Setup Yes We empirically set µ0 = 0, 0 = 1E-5, β0 = 1, 0 = 1E3, T = 200, = 0.25.