Table 16 Results in multiclass hope speech classification with tuned parameters.

From: Analyzing hope speech from psycholinguistic and emotional perspectives

Models

Features

Weighted

Macro

Accuracy

Precision

Recall

F1

Precision

Recall

F1

LightGBM

N-grams

0.6412

0.6596

0.6448

0.5715

0.5213

0.5368

0.6596

LIWC Emotions Sentiments

0.6563

0.6747

0.6579

0.5968

0.5312

0.5507

0.6747

LIWC Emotions Sentiments N-grams

0.6772

0.6923

0.6793

0.6181

0.5613

0.5800

0.6923

CatBoost

N-grams

0.6407

0.6620

0.6427

0.5811

0.5161

0.5350

0.6620

LIWC Emotions Sentiments

0.6558

0.6724

0.6515

0.6096

0.5197

0.5414

0.6724

LIWC Emotions Sentiments N-grams

0.6645

0.6789

0.6586

0.6265

0.5295

0.5539

0.6789

XGBoost

N-grams

0.6334

0.6526

0.6387

0.5582

0.5168

0.5308

0.6526

LIWC Emotions Sentiments

0.6469

0.6650

0.6493

0.5826

0.5234

0.5412

0.6650

LIWC Emotions Sentiments N-grams

0.6706

0.6854

0.6734

0.6078

0.5572

0.5745

0.6854

Back to article page