File size: 2,062 Bytes
a2479cb eb5de39 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 |
---
configs:
- config_name: default
data_files:
- split: test
path: qrels/test.jsonl
- config_name: corpus
data_files:
- split: corpus
path: corpus.jsonl
- config_name: queries
data_files:
- split: queries
path: queries.jsonl
---
## Dataset Summary
**CQADupstack-stats-Fa** is a Persian (Farsi) dataset developed for the **Retrieval** task, focusing on **duplicate question detection** in community question-answering (CQA) platforms. This dataset is a translated version of the **"Cross Validated" (Stats)** subforum from the English **CQADupstack** collection and is part of the **FaMTEB** benchmark under the **BEIR-Fa** suite.
- **Language(s):** Persian (Farsi)
- **Task(s):** Retrieval (Duplicate Question Retrieval)
- **Source:** Translated from English using Google Translate
- **Part of FaMTEB:** Yes — under BEIR-Fa
## Supported Tasks and Leaderboards
This dataset evaluates **text embedding models** on their ability to retrieve **semantically similar or duplicate questions** in domains such as **statistics, machine learning, and data analysis**. Results can be viewed on the **Persian MTEB Leaderboard** (filter by language: Persian).
## Construction
The dataset was generated through:
- Translating the "stats" StackExchange subforum from CQADupstack via **Google Translate API**
- Retaining the original CQA structure for **duplicate retrieval**
According to the FaMTEB paper, the BEIR-Fa datasets underwent quality validation via:
- **BM25 retrieval performance comparisons** (English vs Persian)
- **GEMBA-DA evaluation**, leveraging LLMs for translation quality assurance
## Data Splits
As per the FaMTEB paper (Table 5), all **CQADupstack-Fa** sub-datasets are part of a shared evaluation pool:
- **Train:** 0 samples
- **Dev:** 0 samples
- **Test:** 480,902 samples (aggregate)
The exact subset size for `cqadupstack-stats-fa` is approximately **43.8k examples** (user-provided). For detailed splits, consult the FaMTEB dataset repository or maintainers.
|