Pascal Sun

Pascal Sun

From Momentary Spark to Eternal Intelligence

Sitemap

A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.

Pages

Posts

Full Stack AI Engineer Bootcamp

less than 1 minute read

Published:

Created and delivered UWA’s first Full Stack AI Engineer Bootcamp, combining modern software engineering (Python, React, Django, Docker) with practical AI development. The bootcamp covers 12 modules in 3 weeks, teaching students how to build and deploy production-ready AI applications. The curriculum focuses on real-world implementation, from development setup to deploying AI-enhanced full-stack applications. You can check the details of the bootcamp here: https://tutorial.nlp-tlp.org/.

portfolio

publications

Graph Embeddings for Non-IID Data Feature Representation Learning

Published in Australasian Conference on Data Mining, 2022

Best Research Paper Award at AusDM 2022

Recommended citation: Sun, Q., Liu, W., Huynh, D., Reynolds, M. (2022). Graph Embeddings for Non-IID Data Feature Representation Learning. In: Park, L.A.F., et al. Data Mining. AusDM 2022. Communications in Computer and Information Science, vol 1741. Springer, Singapore. https://doi.org/10.1007/978-981-19-8746-5_4
Download Paper | Download Slides

Are Graph Embeddings the Panacea?

Published in Pacific-Asia Conference on Knowledge Discovery and Data Mining, 2024

An Empirical Survey from the Data Fitness Perspective

Recommended citation: Sun, Q., Huynh, D.Q., Reynolds, M., Liu, W. (2024). Are Graph Embeddings the Panacea?. In: Yang, DN., Xie, X., Tseng, V.S., Pei, J., Huang, JW., Lin, J.CW. (eds) Advances in Knowledge Discovery and Data Mining. PAKDD 2024. Lecture Notes in Computer Science(), vol 14646. Springer, Singapore. https://doi.org/10.1007/978-981-97-2253-2_32
Download Paper | Download Slides

TriagedMSA: Triaging Sentimental Disagreement in Multimodal Sentiment Analysis

Published in IEEE Transactions on Affective Computing, 2025

Existing multimodal sentiment analysis models are effective at capturing sentiment commonalities across different modalities and discerning emotions. However, these models still face significant challenges when analyzing samples with sentiment polarity differences across modalities. Neural networks struggle to process such divergent sentiment samples, particularly when they are scarce within datasets. While larger datasets could help address this limitation, collecting and annotating them is resource-intensive. To overcome this challenge, we propose TriagedMSA, a multimodal sentiment analysis model with triage capability. Our model introduces the Sentiment Disagreement Triage Network, which identifies sentiment disagreement between modalities within a sample. This triage mechanism reduces mutual influence by learning to distinguish between samples of sentiment agreement and disagreement. To process these two sample types, we develop the Sentiment Selection Attention Network and the Sentiment Commonality Attention Network, both of which enhance modality interaction learning. Furthermore, we propose the Adaptive Polarity Detection (APD) algorithm, which ensures the generalizability of our model across different datasets, regardless of whether unimodal labels are available. The APD algorithm adaptively determines sentiment polarity disagreement or agreement between modalities. We conduct experiments on three multimodal sentiment analysis datasets: CMU-MOSI, CMU-MOSEI and CH-SIMS.v2. The results demonstrate that our proposed methodology outperforms existing state-of-the-art approaches.

Recommended citation: [TriagedMSA: Triaging Sentimental Disagreement in Multimodal Sentiment Analysis](https://ieeexplore.ieee.org/abstract/document/10819657) (Y. Luo et al.)
Download Paper

Docs2KG: A Human-LLM Collaborative Approach to Unified Knowledge Graph Construction from Heterogeneous Documents

Published in WWW 25: Companion Proceedings of the ACM on Web Conference 2025, 2025

Uncovering Knowledge from Chaos

Recommended citation: Qiang Sun, Yuanyi Luo, Wenxiao Zhang, Sirui Li, Jichunyang Li, Kai Niu, Xiangrui Kong, and Wei Liu. 2025. Docs2KG: A Human-LLM Collaborative Approach to Unified Knowledge Graph Construction from Heterogeneous Documents. In Companion Proceedings of the ACM on Web Conference 2025 (WWW 25). Association for Computing Machinery, New York, NY, USA, 801–804. https://doi.org/10.1145/3701716.3715309
Download Paper | Download Slides

TimelineKGQA: A Comprehensive Question-Answer Pair Generator for Temporal Knowledge Graphs

Published in WWW 25: Companion Proceedings of the ACM on Web Conference 2025, 2025

TimelineKGQA, a universal temporal QA generator applicable to any TKGs

Recommended citation: Qiang Sun, Sirui Li, Du Huynh, Mark Reynolds, and Wei Liu. 2025. TimelineKGQA: A Comprehensive Question-Answer Pair Generator for Temporal Knowledge Graphs. In Companion Proceedings of the ACM on Web Conference 2025 (WWW 25). Association for Computing Machinery, New York, NY, USA, 797–800. https://doi.org/10.1145/3701716.3715308
Download Paper | Download Slides

DAG-Think-Twice: Causal Structure Guided Elicitation of Causal Reasoning in LLMs

Published in Pacific-Asia Conference on Knowledge Discovery and Data Mining, 2025

Reasoning based way to do the finetuning

Recommended citation: Deng, Z., Sun, Q., Li, J., Liu, W. (2025). DAG-Think-Twice: Causal Structure Guided Elicitation of Causal Reasoning in LLMs. In: Wu, X., et al. Data Science: Foundations and Applications. PAKDD 2025. Lecture Notes in Computer Science(), vol 15876. Springer, Singapore. https://doi.org/10.1007/978-981-96-8298-0_33
Download Paper | Download Slides

talks

teaching

Teaching experience 1

Undergraduate course, University 1, Department, 2014

This is a description of a teaching experience. You can use markdown like any other post.

Teaching experience 2

Workshop, University 1, Department, 2015

This is a description of a teaching experience. You can use markdown like any other post.