Update README.md
Browse files
README.md
CHANGED
|
@@ -29,16 +29,20 @@ The dataset captures diverse linguistic styles, authentic code-mixing, and real-
|
|
| 29 |
The dataset includes four major text categories, representing different levels of English–Bengali mixing.
|
| 30 |
Each group is evenly structured across the dataset for balanced linguistic coverage.
|
| 31 |
|
| 32 |
-
🇬🇧 English Texts : 15,000 ▓▓▓▓▓░░░░░░░░░░░░░░░░░ 15%
|
| 33 |
|
| 34 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 35 |
|
| 36 |
-
🌍 English–Bengali Mixed Texts : 35,000 ▓▓▓▓▓▓▓▓▓▓▓▓░░░░░░░░░░ 35%
|
| 37 |
|
| 38 |
-
🔤 English–Roman Bengali Mixed : 35,000 ▓▓▓▓▓▓▓▓▓▓▓▓░░░░░░░░░░ 35%
|
| 39 |
-
---------------------------------------------------------------
|
| 40 |
|
| 41 |
-
🧮 **Total Samples** : 100,000
|
| 42 |
|
| 43 |
|
| 44 |
📘 This distribution ensures the dataset provides a rich combination of monolingual and code-mixed samples suitable for multilingual model training.
|
|
@@ -48,11 +52,14 @@ Each group is evenly structured across the dataset for balanced linguistic cover
|
|
| 48 |
The dataset exhibits extensive lexical diversity across three language layers.
|
| 49 |
Word counts demonstrate the dominance of English syntax, with embedded Bengali and Roman Bengali expressions adding cultural and emotional context.
|
| 50 |
|
| 51 |
-
✴️ English Words
|
| 52 |
-
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
|
|
|
|
|
|
|
|
|
| 56 |
|
| 57 |
|
| 58 |
📙 The mix of three distinct scripts provides valuable linguistic variability, allowing models to learn fine-grained lexical and orthographic distinctions.
|
|
@@ -62,10 +69,12 @@ Word counts demonstrate the dominance of English syntax, with embedded Bengali a
|
|
| 62 |
The dataset follows a binary sentiment structure with positive and negative review labels.
|
| 63 |
Unlike a perfectly balanced dataset, it naturally reflects real-world customer behavior — where users share more positive experiences.
|
| 64 |
|
| 65 |
-
😊 Positive Reviews
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
|
|
|
|
|
|
| 69 |
|
| 70 |
|
| 71 |
📗 This realistic sentiment imbalance provides a natural testing ground for building sentiment classification models robust to class skew.
|
|
|
|
| 29 |
The dataset includes four major text categories, representing different levels of English–Bengali mixing.
|
| 30 |
Each group is evenly structured across the dataset for balanced linguistic coverage.
|
| 31 |
|
|
|
|
| 32 |
|
| 33 |
+
🇬🇧 English Texts: 15,000 ▓▓▓▓▓░░░░░░░░░░░░░░░░░ 15%
|
| 34 |
+
|
| 35 |
+
🇧🇩 Bengali Texts: 15,000 ▓▓▓▓▓░░░░░░░░░░░░░░░░░ 15%
|
| 36 |
+
|
| 37 |
+
🌍 English–Bengali Mixed Texts: 35,000 ▓▓▓▓▓▓▓▓▓▓▓▓░░░░░░░░░░ 35%
|
| 38 |
+
|
| 39 |
+
🔤 English–Roman Bengali Mixed: 35,000 ▓▓▓▓▓▓▓▓▓▓▓▓░░░░░░░░░░ 35%
|
| 40 |
+
|
| 41 |
+
|
| 42 |
+
🧮 Total Samples: 100,000
|
| 43 |
|
|
|
|
| 44 |
|
|
|
|
|
|
|
| 45 |
|
|
|
|
| 46 |
|
| 47 |
|
| 48 |
📘 This distribution ensures the dataset provides a rich combination of monolingual and code-mixed samples suitable for multilingual model training.
|
|
|
|
| 52 |
The dataset exhibits extensive lexical diversity across three language layers.
|
| 53 |
Word counts demonstrate the dominance of English syntax, with embedded Bengali and Roman Bengali expressions adding cultural and emotional context.
|
| 54 |
|
| 55 |
+
✴️ English Words: 6,870,500 ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 71.5%
|
| 56 |
+
|
| 57 |
+
🔡 Bengali Words: 2,136,460 ▓▓▓▓▓▓▓▓░░░░░░░░░░░░░░░░░░ 22.2%
|
| 58 |
+
|
| 59 |
+
🔠 Roman Bengali Words: 601,220 ▓▓░░░░░░░░░░░░░░░░░░░░░░░░ 6.3%
|
| 60 |
+
|
| 61 |
+
🧾 Total Word Count: 9,608,180
|
| 62 |
+
|
| 63 |
|
| 64 |
|
| 65 |
📙 The mix of three distinct scripts provides valuable linguistic variability, allowing models to learn fine-grained lexical and orthographic distinctions.
|
|
|
|
| 69 |
The dataset follows a binary sentiment structure with positive and negative review labels.
|
| 70 |
Unlike a perfectly balanced dataset, it naturally reflects real-world customer behavior — where users share more positive experiences.
|
| 71 |
|
| 72 |
+
😊 Positive Reviews: 79.3% ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓░░ 79,300
|
| 73 |
+
|
| 74 |
+
😠 Negative Reviews: 20.7% ▓▓▓▓░░░░░░░░░░░░░░░░░░░░░░ 20,700
|
| 75 |
+
|
| 76 |
+
💡 Total Samples: 100,000
|
| 77 |
+
|
| 78 |
|
| 79 |
|
| 80 |
📗 This realistic sentiment imbalance provides a natural testing ground for building sentiment classification models robust to class skew.
|