Proposed model scores and sentiment prediction. حسب Xiaohan Yu (7525940)

الموضوعات:

This package of Python scripts implements the proposed unsupervised sentiment analysis approach. It contains a pre-processing script (preprocess.py), a script that calculates BERT... حسب Xiaohan Yu (7525940)

الموضوعات:

JSON files with the zero-shot sentiment classification predictions/outputs by ChatGPT-4o and DeepSeek r1. These outputs were directly compared with our model. حسب Xiaohan Yu (7525940)

الموضوعات:

The proposed model compares with state-of-the-art (SOTA) methods. حسب Xiaohan Yu (7525940)

الموضوعات:

Comparison of the proposed model with existing models. حسب Xiaohan Yu (7525940)

الموضوعات:

Comparison of the proposed model with ChatGPT-4o and DeepSeek R1. حسب Xiaohan Yu (7525940)

الموضوعات:

BERT Encoder for contextualised embeddings. حسب Xiaohan Yu (7525940)

الموضوعات:

A JSON file with ground truth sentiment labels used in evaluation and comparison to LLM prediction. حسب Xiaohan Yu (7525940)

الموضوعات:

Architecture of the GAT model. حسب Xiaohan Yu (7525940)

الموضوعات:

There are Python notebooks and scripts to build a zero-shot sentiment classifier using ChatGPT-4o and DeepSeek r1. These codes replicate the comparative evaluation experiments docu... حسب Xiaohan Yu (7525940)

الموضوعات:

Classical Chinese Texts Corpus: This corpus/repository includes the classical Chinese literature used in the experiments of sentiment analysis presented in this paper. It contains... حسب Xiaohan Yu (7525940)

الموضوعات:

Performance on multi-relation graphs using both friendship and follower relations. Boldface letters indicate the best results. حسب Yan Li (23143)

الموضوعات:

The accuracy of G-CutMix and its variants. Boldface letters are used to mark the best results. حسب Yan Li (23143)

الموضوعات:

Node Cutmix module in our proposed G-CutMix. حسب Yan Li (23143)

الموضوعات:

Performance of bot detection using followers relationship graph. Acc and F1 stands for accuracy and F1-score (macro), respectively. Boldface letters are used to mark the best resul... حسب Yan Li (23143)

الموضوعات:

Accuracy of different training sets using different relationship graph. حسب Yan Li (23143)

الموضوعات:

Computational complexity comparisons for different methods. حسب Yan Li (23143)

الموضوعات:

Statistics of bot detection datasets used in the paper. حسب Yan Li (23143)

الموضوعات:

Performance of bot detection using friends relationship graph. Acc and F1 stands for accuracy and F1-score (macro), respectively. Boldface letters are used to mark the best results... حسب Yan Li (23143)

الموضوعات:

Performance of G-CutMix with different <i>α.</i> حسب Yan Li (23143)

الموضوعات: