MCINext/cqadupstack-tex-fa

1---2configs:3  - config_name: default4    data_files:5      - split: test6        path: qrels/test.jsonl7  - config_name: corpus8    data_files:9      - split: corpus10        path: corpus.jsonl11  - config_name: queries12    data_files:13      - split: queries14        path: queries.jsonl15---16## Dataset Summary17 18**CQADupstack-tex-Fa** is a Persian (Farsi) dataset curated for the **Retrieval** task, specifically targeting **duplicate question detection** in community question-answering (CQA) forums. This dataset is a translation of the **"TeX - LaTeX" StackExchange subforum** from the English **CQADupstack** collection and is part of the **FaMTEB** benchmark under the **BEIR-Fa** suite.19 20- **Language(s):** Persian (Farsi)  21- **Task(s):** Retrieval (Duplicate Question Retrieval)  22- **Source:** Translated from English using Google Translate  23- **Part of FaMTEB:** Yes — under BEIR-Fa24 25## Supported Tasks and Leaderboards26 27This dataset assesses the ability of **text embedding models** to identify **semantically equivalent or duplicate questions** within the domain of **TeX and LaTeX programming**. Evaluation results appear on the **Persian MTEB Leaderboard** (filter by language: Persian).28 29## Construction30 31The construction process included:32 33- Translating the **"TeX - LaTeX"** subforum from the CQADupstack dataset using the **Google Translate API**34- Retaining the original duplicate-retrieval task structure35 36The **FaMTEB** paper confirms quality through:37 38- **BM25 retrieval comparisons** across English and Persian versions  39- **GEMBA-DA framework**, which employed LLMs to verify translation fidelity40 41## Data Splits42 43Per FaMTEB paper (Table 5), all **CQADupstack-Fa** sub-datasets share the same test pool:44 45- **Train:** 0 samples  46- **Dev:** 0 samples  47- **Test:** 480,902 samples (aggregate)48 49The individual split count for `cqadupstack-tex-fa` is not provided separately. For detailed partitioning, refer to FaMTEB resources.50 51**Total (user-reported):** ~76.2k examples52