docs: add concise README for the text sentiment example

Author: IamSavitha

examples/README.MD

@@ -0,0 +1,70 @@
+# Text Sentiment (SVM) with Class Rebalancing — imbalanced-learn Example
+
+**What:** A small, runnable demo for 3-class sentiment (negative / neutral / positive) using:
+
+```
+TF-IDF → RandomUnderSampler → LinearSVC
+```
+
+**Why:** Text features are sparse (TF-IDF). Oversampling methods like SMOTE target dense/continuous data;
+
+**under-sampling** works out-of-the-box for sparse text.
+
+---
+
+## Files
+
+* `examples/text_sentiment_svm_with_resampling.py` — example script (CLI)
+* `imblearn/tests/test_text_sentiment_example.py` & `..._cli.py` — fast smoke and unit tests
+
+---
+
+## Setup
+
+```bash
+# in a virtual env
+pip install -e .          # install this repo
+pip install datasets matplotlib pytest
+# optional: keep dataset cache local
+export HF_DATASETS_CACHE="$PWD/.hf_cache"
+```
+
+## Run
+
+```bash
+python examples/text_sentiment_svm_with_resampling.py --plot --max-samples 6000
+```
+
+**Outputs**
+
+* Prints **balanced accuracy** + **classification report**
+* Saves `confmat_svm_imblearn.png` when `--plot` is used
+
+**CLI options**
+
+```
+--max-samples  INT   Limit training size (None = full). Default: 6000
+--plot               Save confusion matrix image
+--output PATH        Image path (default: confmat_svm_imblearn.png)
+```
+
+---
+
+## Tests
+
+```bash
+pytest -q imblearn/tests/test_text_sentiment_example.py
+pytest -q imblearn/tests/test_text_sentiment_example_cli.py
+```
+
+Tests are quick, deterministic, and skipped if `datasets` isn’t installed.
+
+---
+
+## Notes
+
+* Metric focus: **balanced accuracy** & **macro-F1** (better for imbalance)
+* Reproducible: fixed `random_state`, controllable `--max-samples`
+* Troubleshooting: low disk? use `pip --no-cache-dir`, clear caches, keep only one env active
+
+---