Papers by Tianze Shi

16 papers
Transition-based Bubble Parsing: Improvements on Coordination Structure Prediction (2021.acl-long)

Copied to clipboard

Challenge: Existing bubble representations encoding coordination boundaries and internal relationships are difficult to detect and parse .
Approach: They propose a bubble parser to perform coordination structure identification and dependency-based syntactic analysis simultaneously.
Outcome: The proposed bubble parser beats state-of-the-art approaches on coordination structure prediction . the proposed system is based on a GENIA corpus and a Penn treebank .
Improving Coverage and Runtime Complexity for Exact Inference in Non-Projective Transition-Based Dependency Parsers (N18-2)

Copied to clipboard

Challenge: Non-projective dependency trees account for 12.59% of all training sentences in the annotated Universal Dependencies (UD) 2.1 data.
Approach: They generalize Cohen et al.'s (2011) parser to a family of non-projective transition-based dependency parsers allowing polynomial-time exact inference.
Outcome: The proposed system can be extended to include a variant that reduces time complexity to O(n6), improving over the known bounds in exact inference for non-projective transition-based parsing.
Weakly Supervised Headline Dependency Parsing (2022.findings-emnlp)

Copied to clipboard

Challenge: English news headlines have unique syntactic properties documented in linguistics literature since the 1930s.
Approach: They propose to provide the first news headline corpus of annotated syntactic dependency trees to evaluate existing NLP parsers on news headlines.
Outcome: The proposed method improves performance across different news outlets, but is moderated by constructions idiosyncratic to outlet.
Global Transition-based Non-projective Dependency Parsing (P18-1)

Copied to clipboard

Challenge: Until recently, transition-based dependency parsers were limited to approximate inference due to their incompatibility with rich feature models.
Approach: They propose a transition-based parser with high coverage on non-projective treebanks to support non- projective parsing.
Outcome: The proposed approach is more efficient than its projective counterpart in non-projective languages.
AgentThink: A Unified Framework for Tool-Augmented Chain-of-Thought Reasoning in Vision-Language Models for Autonomous Driving (2025.findings-emnlp)

Copied to clipboard

Challenge: Vision-Language Models struggle with hallucinations, inefficient reasoning, and limited real-world validation hinders accurate perception and robust step-by-step reasoning.
Approach: AgentThink integrates Chain-of-Thought reasoning with dynamic, agent-style tool invocation for autonomous driving tasks.
Outcome: Experiments on the DriveLMM-o1 benchmark show AgentThink significantly boosts overall reasoning scores by 53.91% and enhances answer accuracy by 33.54% .
Evaluating the Impact of Model Scale for Compositional Generalization in Semantic Parsing (2022.emnlp-main)

Copied to clipboard

Challenge: Pre-trained language models struggle on out-of-distribution compositional generalization . recent work shows considerable improvements on many NLP tasks from model scaling .
Approach: They evaluate encoder-decoder models up to 11B parameters and decoder-only models up 540B parameters . they compare scaling curves for fine-tuning, prompt tuning, and in-context learning methods .
Outcome: The proposed scaling methods improve compositional generalization on many tasks . fine-tuning generally has flat or negative scaling curves on out-of-distribution compositional . larger models are better at modeling the syntax of the output space, the study finds .
Semantic Role Labeling as Syntactic Dependency Parsing (2020.emnlp-main)

Copied to clipboard

Challenge: Using propBank-style semantic role labeling, we reduce the task to syntactic dependency parsing.
Approach: They propose to convert SRL annotations into dependency tree representations through joint labels that permit highly accurate recovery back to the original format.
Outcome: The proposed scheme reduces the task of (span-based) PropBank-style semantic role labeling to syntactic dependency parsing.
Learning Syntax from Naturally-Occurring Bracketings (2021.naacl-main)

Copied to clipboard

Challenge: a new method for learning naturally-occurring bracketings is developed . it uses noisy and incomplete data to induce syntactic structures .
Approach: They propose a partial-brackets-aware structured ramp loss in learning to address this challenge . they show that distantly-supervised models trained on naturally-occurring bracketing data are more accurate . constituency is a foundational building block for phrase-structure grammars, they argue .
Outcome: The proposed model achieves an unlabeled F1 score for constituency parsing on the English WSJ corpus.
Editing-Based SQL Query Generation for Cross-Domain Context-Dependent Questions (D19-1)

Copied to clipboard

Challenge: Generating SQL queries from user utterances is an important task to help end users acquire information from databases.
Approach: They propose a context-dependent text-to-SQL generation task that edits previous queries . they use an utterance-table encoder and a table-aware decoder to incorporate context .
Outcome: The proposed model is flexible to change individual tokens and robust to error propagation.
Valency-Augmented Dependency Parsing (D18-1)

Copied to clipboard

Challenge: valency analysis is a complex task that requires a large number of subcategorizations, such as the number and types of syntactic dependents.
Approach: They propose a parsing approach that explicitly models the number and types of syntactic dependents as valency patterns and a probabilistic model for tagging them.
Outcome: The proposed approach outperforms the state-of-the-art labeled attachment score on 53 treebanks representing 41 languages and outperformed the previous state- of-the art labeles by 0.7.
Diversity-Aware Batch Active Learning for Dependency Parsing (2021.naacl-main)

Copied to clipboard

Challenge: a high annotation cost for dependency parsers is a challenge . batch active learning (AL) is based on batch mode, which is more efficient for annotators to label in bulk.
Approach: They propose to reduce the number of labeled examples needed to train a strong dependency parser using batch active learning.
Outcome: The proposed approach improves on an English newswire corpus by enforcing diversity in the sampled batches.
On the Potential of Lexico-logical Alignments for Semantic Parsing to SQL Queries (2020.findings-emnlp)

Copied to clipboard

Challenge: Large-scale semantic parsing datasets annotated with logical forms have enabled advances in supervised approaches.
Approach: They propose to enrich English-language questions with SQL equivalents and alignments . they propose to use supervised attention and an auxiliary objective to disambiguate references .
Outcome: The proposed method improves over strong baselines by 4.4% execution accuracy.
Extracting Headless MWEs from Dependency Parse Trees: Parsing, Tagging, and Joint Modeling Approaches (2020.acl-main)

Copied to clipboard

Challenge: Headless multi-word expressions are frequent in natural language but lack internal syntactic dominance relations.
Approach: They propose an efficient joint decoding algorithm that combines scores from both strategies.
Outcome: The proposed algorithm combines scores from parsing and tagging for predicting flat MWEs . the proposed algorithm is more accurate than parse and more efficient for non-BERT features .
CoSQL: A Conversational Text-to-SQL Challenge Towards Cross-Domain Natural Language Interfaces to Databases (D19-1)

Copied to clipboard

Challenge: CoSQL is a corpus for building cross-domain, general-purpose database querying dialogue systems.
Approach: They present a corpus for building cross-domain, general-purpose database querying dialogue systems . they use a Wizard-of-Oz collection of 3k turns plus 10k+ annotated SQL queries .
Outcome: The proposed corpus is based on a Wizard-of-Oz dataset of 3k dialogues querying 200 complex DBs spanning 138 domains.
AgencyBench: Benchmarking the Frontiers of Autonomous Agents in 1M-Token Real-World Contexts (2026.acl-long)

Copied to clipboard

Challenge: Existing benchmarks focus on single agentic capability, failing to capture long-horizon real-world scenarios.
Approach: They propose a benchmark that evaluates 6 agentic capabilities across 32 real-world scenarios.
Outcome: Experiments show that closed-source models outperform open-source model (48.4% vs 32.1%) integrating models with advanced scaffolds to form autonomous agents is a paradigm shift.
Optimizing Pre-Training Data Mixtures with Mixtures of Data Expert Models (2025.acl-long)

Copied to clipboard

Challenge: Existing methods to optimize language model pre-training data mixtures are difficult due to the complexity of the data mixture.
Approach: They propose a method to optimize language model pre-training data mixtures by approximating cross-entropy loss via a Mixture of Data Experts (MDE).
Outcome: The proposed method improves performance on a slimPajama dataset with a mixture of data experts.

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