NeoCyber/m-e5-small-vlsp2018-hotel

Aspect-based sentiment model exported from vlsp-2018-hotel-e5-small-best.pt.

Training Metadata

• Base model: intfloat/multilingual-e5-small
• Epoch: 19
• Multi branch: False
• Aspect count: 34
• Sentiment labels: none, positive, negative, neutral

Training Dataset

• Dataset: VLSP 2018 Sentiment Analysis - Hotel
• Local dataset path: training/datasets/vlsp2018_hotel
• Result source: training/pipeline/full_pipeline_hotel.ipynb

Test Metrics (Weighted Avg)

Report	Precision	Recall	F1	Support
aspect_category	0.937400	0.941520	0.931863	20400
aspect_category_polarity	0.931039	0.933578	0.919990	20400

Checkpoint Metrics

{
  "loss": 0.1830130910190443,
  "accuracy": 0.945485294117647,
  "f1": 0.5439991051406502
}

Aspects

• FACILITIES#CLEANLINESS
• FACILITIES#COMFORT
• FACILITIES#DESIGN&FEATURES
• FACILITIES#GENERAL
• FACILITIES#MISCELLANEOUS
• FACILITIES#PRICES
• FACILITIES#QUALITY
• FOOD&DRINKS#MISCELLANEOUS
• FOOD&DRINKS#PRICES
• FOOD&DRINKS#QUALITY
• FOOD&DRINKS#STYLE&OPTIONS
• HOTEL#CLEANLINESS
• HOTEL#COMFORT
• HOTEL#DESIGN&FEATURES
• HOTEL#GENERAL
• HOTEL#MISCELLANEOUS
• HOTEL#PRICES
• HOTEL#QUALITY
• LOCATION#GENERAL
• ROOMS#CLEANLINESS
• ROOMS#COMFORT
• ROOMS#DESIGN&FEATURES
• ROOMS#GENERAL
• ROOMS#MISCELLANEOUS
• ROOMS#PRICES
• ROOMS#QUALITY
• ROOM_AMENITIES#CLEANLINESS
• ROOM_AMENITIES#COMFORT
• ROOM_AMENITIES#DESIGN&FEATURES
• ROOM_AMENITIES#GENERAL
• ROOM_AMENITIES#MISCELLANEOUS
• ROOM_AMENITIES#PRICES
• ROOM_AMENITIES#QUALITY
• SERVICE#GENERAL