Answer Complex Questions: Path Ranker Is All You Need

By Ke Zhan, Enrui Hu, Chengzhen Fu, Lan Luo, Hao Jiang, Yantao Jia, Fan Yu, Zhicheng Dou, Zhao Cao, Lei Chen. 2021

32 citations recorded by the source corpus.

Currently, the most popular method for open-domain Question Answering (QA) adopts "Retriever and Reader" pipeline, where the retriever extracts a list of candidate documents from a large set of documents followed by a ranker to rank the most relevant documents and the reader extracts answer from the candidates. Existing studies take the greedy strategy in the sense that they only use samples for ranking at the current hop, and ignore the global information across the whole documents. In this paper, we propose a purely rank-based framework Thinking Path Re-Ranker (TPRR), which is comprised of Thinking Path Ranker (TPR) for generating document sequences called "a path" and External Path Reranker (EPR) for selecting the best path from candidate paths generated by TPR. Specifically, TPR leverages the scores of a dense model and conditional probabilities to score the full paths. Moreover, to further enhance the performance of the dense ranker in the iterative training, we propose a "thinking" negatives selection method that the top-K candidates treated as negatives in the current hop are adjusted dynamically through supervised signals. After achieving multiple supporting paths through TPR, the EPR component which integrates several fine-grained training tasks for QA is used to select the best path for answer extraction. We have tested our proposed solution on the multi-hop dataset "HotpotQA" with a full wiki set ting, and the results show that TPRR significantly outperforms the existing state-of-the-art models. Moreover, our method has won the first place in the HotpotQA official leaderboard since Feb 1, 2021 under the Fullwiki setting. Code is available at https://gitee.com/mindspore/mindspore/ tree/master/model_zoo/research/nlp/tprr.

Subfields: Artificial Intelligence, Information Systems, Information Retrieval & Search, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Chenliang Li
2. Xueqi Cheng
3. Songfang Huang
4. Bin Bi
5. Yanyan Lan
6. Ming Yan
7. Liang Pang
8. Wei Wang

In Information Systems

1. Fei Huang
2. Xueqi Cheng
3. Ming Yan
4. Luo Si
5. Songfang Huang
6. Huawei Shen
7. Yanyan Lan
8. Wei Wang

In Information Retrieval & Search

1. Yunchang Zhu
2. Xueqi Cheng
3. Ming Yan
4. Songfang Huang
5. Huawei Shen
6. Yanyan Lan
7. Wei Wang
8. Liang Pang

In Natural Language Processing

1. Xueqi Cheng
2. Liang Pang
3. Yanyan Lan
4. Fei Huang
5. Chenliang Li
6. Haiyang Xu
7. Yunchang Zhu
8. Ming Yan

Citations in the corpus

• Adaptive Information Seeking for Open-Domain Question Answering (2021, methodology): “Notably, on the HotpotQA fullwiki blind test set7, AISOlarge significantly outperforms the second place TPRR (Xinyu et al., 2021) by 2.”
• Achieving Human Parity on Visual Question Answering (2021, background): “Recently, open-domain QA [53, 58, 78, 96, 97] has become one of the most popular topics in question answering.”