Papers by Zhiwen Xie

5 papers
EmoS: A High-Fidelity Multimodal Benchmark for Fine-grained Streaming Emotional Understanding (2026.acl-long)

Copied to clipboard

Challenge: Existing benchmarks fail to achieve ecological validity, signal clarity, and reliable fine-grained labeling in multimodal Emotion Recognition (MER) Existing datasets lack spontaneity of real-life interactions, resulting in poor quality and inconsistent data quality.
Approach: They propose a bilingual benchmark to resolve limitations of ecological validity and noise in existing datasets by combining strictly filtered static slices with a dynamic Streaming Monologue subset.
Outcome: EmoS provides trusted ground truth that captures continuous emotional evolution.
MAKI: Multi-layer Aligned Knowledge Injection for Structure-aware Knowledge Graph Completion with Large Language Models (2026.findings-acl)

Copied to clipboard

Challenge: Existing knowledge graph completion methods struggle to capture structural information in knowledge graphs (KGs) Existing approaches for KGC focus on learning representations of entities and relations through observed structural patterns.
Approach: They propose a multi-layer Aligned Knowledge Injection model that tightly integrates structured KG information into LLMs through multi-layered alignment.
Outcome: The proposed method outperforms state-of-the-art methods on benchmark datasets.
ReInceptionE: Relation-Aware Inception Network with Joint Local-Global Structural Information for Knowledge Graph Embedding (2020.acl-main)

Copied to clipboard

Challenge: Existing knowledge graph embedding methods are limited in their expressiveness and lack structural information in the embeddable space.
Approach: They propose to use a relation-aware network to learn query embedding . they first explore the Inception network to further increase interactions between head and relation embedders .
Outcome: The proposed network improves performance on WN18RR and FB15k-237 datasets.
Learning Query Adaptive Anchor Representation for Inductive Relation Prediction (2023.findings-acl)

Copied to clipboard

Challenge: Existing methods to infer the missing links between entities are limited to the transductive setting . Query Adaptive Anchor Representation (QAAR) model is based on entity-independent features .
Approach: They propose a query adaptive anchor representation model which extracts one opening subgraph and performs reasoning by one time for all candidate triples.
Outcome: The proposed model outperforms state-of-the-art models in relation prediction task.
A Contextual Alignment Enhanced Cross Graph Attention Network for Cross-lingual Entity Alignment (2020.coling-main)

Copied to clipboard

Challenge: Existing methods for cross-lingual entity alignment ignore useful pre-aligned links between two KGs.
Approach: They propose a novel method that jointly learns embeddings in different KGs by propagating cross-KG information through pre-aligned seed alignments.
Outcome: The proposed method achieves remarkable performance gains on three benchmark cross-lingual entity alignment datasets.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations