This is the repository for our EMNLP 2023 Findings paper:

Large Language Models Know Your Contextual Search Intent: A Prompting Framework for Conversational Search

We conduct experiemnts in Python 3.8.13.

Main packages:

• Generating rewrites:

  • openai==0.26.4
  • numpy==1.23.5

• For evaluation:

  • torch==1.12.0+cu113
  • trec-car-tools==2.6
  • faiss-gpu
  • pyserini==0.17.0

conda create -n llmcs python=3.8
source activate llmcs
pip install -r requirements.txt

1. We provide the preprocessed cast19,20,21 datasets in the datasets folder.

2. demonstrations.json contains four exemplars randomly sampled from the CAsT-22 dataset. We manually write CoT for all of its turns.

LLM4CS contains three prompting methods, including Rewriting Prompt (REW), Rewriting-Then-Response Prompt (RTR), and Rewriting-And-Response Prompt (RAR). We also design chain-of-thought tailored to conversational search intent understanding that can be incorporated into these prompting methods.

To get started, first, you should set your OpenAI API key in generator.py

# TODO: Write your OpenAI API here.
OPENAI_KEYS = [
    'Your key',
]

To perform REW prompting, run:

bash scripts/run_chat_prompt_rewrite.sh

Also, you can enable CoT by running:

bash scripts/run_chat_prompt_cot_rewrite.sh

Similarly, to perform RTR prompting, run:

bash scripts/run_chat_prompt_rewrite_then_response.sh

Note that you need to provide a pre-generated rewrite file into the field of rewrite_file_path for running prompt_rewrite_then_response.py. To enable CoT for RAR, you can set rewrite_file_path to the rewrite file generated using CoT.

--rewrite_file_path="./results/cast20/REW/rewrites.jsonl" \
# --rewrite_file_path="./results/cast20/COT-REW/rewrites.jsonl" \ +COT

Similarly, to perform RAR prompting, run:

bash scripts/run_chat_prompt_rewrite_and_response.sh

Also, you can enable CoT by running:

bash scripts/run_chat_prompt_cot_rewrite_and_response.sh

A rewrites.jsonl file, which contains the rewrites or/and hypothetical responses, will be generated into the work_dir that you set in the running script.

We have provided our generated rewrites.jsonl files in the results folder.

The Keys of rewrites.jsonl:

• predicted_rewrite: rewrite (list)
• preidcted_response: hypothetical response (list)
• other auxiliary information

Note:

We provide some results of REW on CAsT-19 in results/cast19/REW and results/cast19/COT+REW for Case Study mentioned in Appendix B.1. We also provide a part of human evaluation results on CAsT-21 in results/cast21_human_evaluation.part.json for Case Study mentioned in Appendix B.2.

All results will be provided upon acceptance.

We design three aggregation methods, including MaxProb, Mean, and SC, to get the final search intent vector. Then we perform dense retrieval with ANCE (click to download) for evaluation.

To perform evaluation, you should first build the dense index that contains all candidate passage embeddings. There have been many good public repositories that instruct how to build this index, such as AutoRewriter, ConvDR, and ConvTrans. One can refer them to build the index. All passage embeddings generated from ANCE are around 103GB.

Then, run the following script for evaluation:

cd evaluation
bash run_eval_dense_retrieval.sh

We have annotated the important arguments of run_eval_dense_retrieval.sh as below:

# eval_dense_retrieval.sh
# An example of evaluating RTR on cast20.
eval_field_name="predicted_rewrite"
work_dir="../results/cast20/RTR"    # set your the folder containing your `rewrites.jsonl`file

eval_file_path="$work_dir/rewrites.jsonl" \
index_path="" # set the pre-built index which contains all candidate passage emebddings. 
qrel_file_path="../datasets/cast20_qrel.tsv" # set the qrel file path
retrieval_output_path="$work_dir/ance/+q+r+sc" # set your expected output folder path

export CUDA_VISIBLE_DEVICES=0
python eval_dense_retrieval.py \
--eval_file_path=$eval_file_path \
--eval_field_name=$eval_field_name \
--qrel_file_path=$qrel_file_path \
--index_path=$index_path \
--retriever_path="" \ # set the pre-trained ANCE model path.
--use_gpu_in_faiss \
--n_gpu_for_faiss=1 \
--top_n=1000 \
--rel_threshold=2 \ # set 1 for cast19 and 2 for cast20.
--retrieval_output_path=$retrieval_output_path \
--include_query \
--include_response \ # enable `include_response` if you test RTR or RAR prompting.
--aggregation_method="sc" \ # you can set [`maxprob, mean, sc`]

Evaluation results that contains the following three files will be output into retrieval_output_path that you set.

• metrics.res: all evaluation metrics.
• parameters.txt: parameters record of the evaluation.
• res.trec: detailed TREC-style search results.