Xixin Wu


pdf bib
ConvRGX: Recognition, Generation, and Extraction for Self-trained Conversational Question Answering
Tianhua Zhang | Liping Tang | Wei Fang | Hongyin Luo | Xixin Wu | Helen Meng | James Glass
Proceedings of the Third DialDoc Workshop on Document-grounded Dialogue and Conversational Question Answering

Collecting and constructing human-annotated corpora for training conversational question-answering (CQA) models has recently been shown to be inefficient and costly. To solve this problem, previous works have proposed training QA models with automatically generated QA data. In this work, we extend earlier studies on QA synthesis, and propose an efficient QA data generation algorithm under conversational settings. Our model recognizes potential dialogue topics, generates corresponding questions, and extracts answers from grounding passages. To improve the quality of generated QAs and downstream self-training of CQA models, we propose dropout and agreement-based QA selection methods. We conduct experiments on both data augmentation and domain adaptation settings. Experiments on the QuAC and Doc2Dial tasks show that the proposed method can significantly improve the quality of generated QA data, and also improves the accuracy of self-trained CQA models based on the constructed training corpora.

pdf bib
Search Augmented Instruction Learning
Hongyin Luo | Tianhua Zhang | Yung-Sung Chuang | Yuan Gong | Yoon Kim | Xixin Wu | Helen Meng | James Glass
Findings of the Association for Computational Linguistics: EMNLP 2023

Large language models (LLMs) have been significantly improved by instruction fine-tuning, but still lack transparency and the ability to utilize up-to-date knowledge and information. In this work, we propose search-augmented instruction learning (SAIL), which grounds the language generation and instruction following abilities on complex search results generated by in-house and external search engines. With an instruction tuning corpus, we collect search results for each training case from different search APIs and domains, and construct a new search-grounded training set containing (instruction, grounding information, response) triplets. We then fine-tune the LLaMA-7B model on the constructed training set. Since the collected results contain unrelated and disputing languages, the model needs to learn to ground on trustworthy search results, filter out distracting passages, and generate the target response. The search result-denoising process entails explicit trustworthy information selection and multi-hop reasoning, since the retrieved passages might be informative but not contain the instruction-following answer. Experiments show that the fine-tuned SAIL-7B model has a strong instruction-following ability, and it performs significantly better on transparency-sensitive tasks, including open-ended question answering and fact checking.


pdf bib
Grounded Dialogue Generation with Cross-encoding Re-ranker, Grounding Span Prediction, and Passage Dropout
Kun Li | Tianhua Zhang | Liping Tang | Junan Li | Hongyuan Lu | Xixin Wu | Helen Meng
Proceedings of the Second DialDoc Workshop on Document-grounded Dialogue and Conversational Question Answering

MultiDoc2Dial presents an important challenge on modeling dialogues grounded with multiple documents. This paper proposes a pipeline system of “retrieve, re-rank, and generate”, where each component is individually optimized. This enables the passage re-ranker and response generator to fully exploit training with ground-truth data. Furthermore, we use a deep cross-encoder trained with localized hard negative passages from the retriever. For the response generator, we use grounding span prediction as an auxiliary task to be jointly trained with the main task of response generation. We also adopt a passage dropout and regularization technique to improve response generation performance. Experimental results indicate that the system clearly surpasses the competitive baseline and our team CPII-NLP ranked 1st among the public submissions on ALL four leaderboards based on the sum of F1, SacreBLEU, METEOR and RougeL scores.


pdf bib
Coupling Global and Local Context for Unsupervised Aspect Extraction
Ming Liao | Jing Li | Haisong Zhang | Lingzhi Wang | Xixin Wu | Kam-Fai Wong
Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)

Aspect words, indicating opinion targets, are essential in expressing and understanding human opinions. To identify aspects, most previous efforts focus on using sequence tagging models trained on human-annotated data. This work studies unsupervised aspect extraction and explores how words appear in global context (on sentence level) and local context (conveyed by neighboring words). We propose a novel neural model, capable of coupling global and local representation to discover aspect words. Experimental results on two benchmarks, laptop and restaurant reviews, show that our model significantly outperforms the state-of-the-art models from previous studies evaluated with varying metrics. Analysis on model output show our ability to learn meaningful and coherent aspect representations. We further investigate how words distribute in global and local context, and find that aspect and non-aspect words do exhibit different context, interpreting our superiority in unsupervised aspect extraction.