Generating Synthetic Documents for Cross-Encoder Re-Rankers: A Comparative Study of ChatGPT and Human Experts

May 03, 2023 Β· Declared Dead Β· πŸ› arXiv.org

πŸ‘» CAUSE OF DEATH: Ghosted
No code link whatsoever

"No code URL or promise found in abstract"

Evidence collected by the PWNC Scanner

Authors Arian Askari, Mohammad Aliannejadi, Evangelos Kanoulas, Suzan Verberne arXiv ID 2305.02320 Category cs.IR: Information Retrieval Citations 14 Venue arXiv.org Last Checked 4 months ago
Abstract
We investigate the usefulness of generative Large Language Models (LLMs) in generating training data for cross-encoder re-rankers in a novel direction: generating synthetic documents instead of synthetic queries. We introduce a new dataset, ChatGPT-RetrievalQA, and compare the effectiveness of models fine-tuned on LLM-generated and human-generated data. Data generated with generative LLMs can be used to augment training data, especially in domains with smaller amounts of labeled data. We build ChatGPT-RetrievalQA based on an existing dataset, human ChatGPT Comparison Corpus (HC3), consisting of public question collections with human responses and answers from ChatGPT. We fine-tune a range of cross-encoder re-rankers on either human-generated or ChatGPT-generated data. Our evaluation on MS MARCO DEV, TREC DL'19, and TREC DL'20 demonstrates that cross-encoder re-ranking models trained on ChatGPT responses are statistically significantly more effective zero-shot re-rankers than those trained on human responses. In a supervised setting, the human-trained re-rankers outperform the LLM-trained re-rankers. Our novel findings suggest that generative LLMs have high potential in generating training data for neural retrieval models. Further work is needed to determine the effect of factually wrong information in the generated responses and test our findings' generalizability with open-source LLMs. We release our data, code, and cross-encoders checkpoints for future work.
Community shame:
Not yet rated
Community Contributions

Found the code? Know the venue? Think something is wrong? Let us know!

πŸ“œ Similar Papers

In the same crypt β€” Information Retrieval

Died the same way β€” πŸ‘» Ghosted