Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,118 @@
|
|
| 1 |
---
|
| 2 |
-
license:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
+
license: cc-by-4.0
|
| 3 |
+
annotations_creators:
|
| 4 |
+
- NIST
|
| 5 |
+
task_categories:
|
| 6 |
+
- text-retrieval
|
| 7 |
+
- text-ranking
|
| 8 |
+
language:
|
| 9 |
+
- en
|
| 10 |
+
- zh
|
| 11 |
+
multilinguality:
|
| 12 |
+
- multilingual
|
| 13 |
+
pretty_name: NeuCLIRTech
|
| 14 |
+
size_categories:
|
| 15 |
+
- n<1K
|
| 16 |
+
task_ids:
|
| 17 |
+
- document-retrieval
|
| 18 |
+
configs:
|
| 19 |
+
- config_name: queries
|
| 20 |
+
default: true
|
| 21 |
+
data_files:
|
| 22 |
+
- split: eng
|
| 23 |
+
path: data/eng.tsv
|
| 24 |
+
- split: zho
|
| 25 |
+
path: data/zho.tsv
|
| 26 |
+
format: csv
|
| 27 |
+
sep: "\t"
|
| 28 |
+
header: null
|
| 29 |
+
names: ["id", "query"]
|
| 30 |
+
dataset_info:
|
| 31 |
+
features:
|
| 32 |
+
- name: id
|
| 33 |
+
dtype: string
|
| 34 |
+
- name: query
|
| 35 |
+
dtype: string
|
| 36 |
+
- config_name: qrels
|
| 37 |
+
default: false
|
| 38 |
+
data_files: data/qrels.gains.txt
|
| 39 |
+
format: csv
|
| 40 |
+
sep: " "
|
| 41 |
+
header: null
|
| 42 |
+
names: ["id", "ignore", "docid", "relevance"]
|
| 43 |
+
dataset_info:
|
| 44 |
+
features:
|
| 45 |
+
- name: id
|
| 46 |
+
dtype: string
|
| 47 |
+
- name: ignore
|
| 48 |
+
dtype: string
|
| 49 |
+
- name: docid
|
| 50 |
+
dtype: string
|
| 51 |
+
- name: relevance
|
| 52 |
+
dtype: int
|
| 53 |
---
|
| 54 |
+
|
| 55 |
+
# NeuCLIRTech Topics and Queries
|
| 56 |
+
|
| 57 |
+
NeuCLIRTech is an evaluation benchmark for monolingual and cross-language on technical documents
|
| 58 |
+
|
| 59 |
+
The document collection can be found at [neuclir/csl](https://huggingface.co/datasets/neuclir/csl).
|
| 60 |
+
|
| 61 |
+
## Supporting Tasks and Corresponding Data
|
| 62 |
+
|
| 63 |
+
NeuCLIRBench supports three types of tasks: Chinese monolingual and English-Chinese cross-language retrieval.
|
| 64 |
+
The following specifies the documents, queries, and qrels (labels) that should be used for each task.
|
| 65 |
+
|
| 66 |
+
Please report nDCG@20 for all tasks.
|
| 67 |
+
|
| 68 |
+
*We use `:` to indicate different subset under the dataset.*
|
| 69 |
+
|
| 70 |
+
| Task | Documents | Queries | Qrels |
|
| 71 |
+
| --- | --- | --- | --- |
|
| 72 |
+
| Monolingual | `neuclir/csl:csl` | `zho` split of `neuclir/tech:queries` | `neuclir/tech:qrels` |
|
| 73 |
+
| Cross-Language | `neuclir/csl:csl` | `eng` split of `neuclir/tech:queries` | `neuclir/tech:qrels` |
|
| 74 |
+
|
| 75 |
+
|
| 76 |
+
## Baseline Retrieval Results and Run Files
|
| 77 |
+
We also provide all reported baseline retrieval results in the NeuCLIRBench paper.
|
| 78 |
+
Please refer to the paper for the detailed descriptions of each model.
|
| 79 |
+
|
| 80 |
+

|
| 81 |
+
|
| 82 |
+
|
| 83 |
+
### Run Names
|
| 84 |
+
Please refer to the `./runs` directory in this dataset to find all the runs.
|
| 85 |
+
Files follow the naming scheme of `{run_handle}_{task:mono/clir/mlir}_{lang}.trec`. Please refer to the task section for the details.
|
| 86 |
+
|
| 87 |
+
| Run Handle | Model Type | Model Name |
|
| 88 |
+
|:------------------|:-------------------|:----------------------|
|
| 89 |
+
| bm25 | Lexical | BM25 |
|
| 90 |
+
| bm25dt | Lexical | BM25 w/ DT |
|
| 91 |
+
| bm25qt | Lexical | BM25 w/ QT |
|
| 92 |
+
| milco | Bi-Encoder | MILCO |
|
| 93 |
+
| plaidx | Bi-Encoder | PLAID-X |
|
| 94 |
+
| qwen8b | Bi-Encoder | Qwen3 8B Embed |
|
| 95 |
+
| qwen4b | Bi-Encoder | Qwen3 4B Embed |
|
| 96 |
+
| qwen600m | Bi-Encoder | Qwen3 0.6B Embed |
|
| 97 |
+
| arctic | Bi-Encoder | Arctic-Embed Large v2 |
|
| 98 |
+
| splade | Bi-Encoder | SPLADEv3 |
|
| 99 |
+
| fusion3 | Bi-Encoder | Fusion |
|
| 100 |
+
| repllama | Bi-Encoder | RepLlama |
|
| 101 |
+
| me5large | Bi-Encoder | e5 Large |
|
| 102 |
+
| jinav3 | Bi-Encoder | JinaV3 |
|
| 103 |
+
| bgem3sparse | Bi-Encoder | BGE-M3 Sparse |
|
| 104 |
+
| mt5 | Pointwise Reranker | Mono-mT5XXL |
|
| 105 |
+
| qwen3-0.6b-rerank | Pointwise Reranker | Qwen3 0.6B Rerank |
|
| 106 |
+
| qwen3-4b-rerank | Pointwise Reranker | Qwen3 4B Rerank |
|
| 107 |
+
| qwen3-8b-rerank | Pointwise Reranker | Qwen3 8B Rerank |
|
| 108 |
+
| jina-rerank | Pointwise Reranker | Jina Reranker |
|
| 109 |
+
| searcher-rerank | Pointwise Reranker | SEARCHER Reranker |
|
| 110 |
+
| rank1 | Pointwise Reranker | Rank1 |
|
| 111 |
+
| qwq | Listwise Reranker | Rank-K (QwQ) |
|
| 112 |
+
| rankzephyr | Listwise Reranker | RankZephyr 7B |
|
| 113 |
+
| firstqwen | Listwise Reranker | FIRST Qwen3 8B |
|
| 114 |
+
| rq32b | Listwise Reranker | RankQwen-32B |
|
| 115 |
+
|
| 116 |
+
|
| 117 |
+
## Citation
|
| 118 |
+
TBA
|