Table 1. Summary of literature for informative tweets, infrastructure, and human damage identification.
| Ref | Tasks | Features | Supervised models | ||
|---|---|---|---|---|---|
| IN/DM | IF DM | H DM | |||
| Imran et al. (2014) | ✓ | Customized model | AIDR model | ||
| Cresci et al. (2015) | ✓ | Linguistic and ad-hoc features | SVM | ||
| Nguyen et al. (2017) | ✓ | ✓ | Bag of visual words, VGG-16 | CNN | |
| Rudra et al. (2018) | ✓ | Bag of words, low-level lexical, syntactic features | SVM | ||
| Kumar, Singh & Saumya (2019) | ✓ | TF-IDF, Glove | RF, SVM, KNN, NB, CNN, LSTM, GRU | ||
| Li et al. (2019) | ✓ | VGG-19 | Domain-Adversarial Neural Network | ||
| Priya et al. (2020) | ✓ | Latent Dirichlet allocation, Part of speech | Information retrieval | ||
| Alam, Ofli & Imran (2020) | ✓ | Bag of words, sentiment, Spatial features | RF | ||
| Imran et al. (2020) | ✓ | ✓ | Image features | DNN | |
| Alam et al. (2021) | ✓ | FastText | CNN | ||
| Madichetty & Sridevi (2021) | ✓ | ✓ | ✓ | Low-level lexical, syntactic, and frequency features | SVM, RF, GB, AB, Bagging |
| Krishna, Srinivas & Prasad Reddy (2022) | ✓ | TF-IDF, word2vec, Glove | Majority-voting ensemble model | ||
| Imran et al. (2022) | ✓ | —— | Computer vision models | ||
| Koshy & Elango (2023) | ✓ | Roberta, biLSTM | biLSTM | ||
| Paul, Sahoo & Balabantaray (2023) | ✓ | ✓ | Word embeddings, bag of words | CNN, GRU, SkipCNN | |
| Alam et al. (2023) | ✓ | ✓ | ResNet18, VGG16 | Deep learning models | |
| Lv, Wang & Shao (2023) | ✓ | ResNet50, Word embeddings | Bi-GRU | ||
| Garg, Chakraborty & Dandapat (2023) | ✓ | ✓ | —— | Ontology-based approach | |
| Asinthara, Jayan & Jacob (2023) | ✓ | ✓ | TF-IDF, word2vec | SVM, bi-LSTM | |
| Arathi & Sasikala (2023) | ✓ | Glove model, metadata features | RF | ||
| Dasari, Gorla & Prasad Reddy (2023) | ✓ | TF-IDF, word2vec, Glove | Stacking ensemble model | ||
| Giri & Deepak (2023) | ✓ | Image features, word embeddings | Ontology infused DL model | ||
| Madichetty & Madisetty (2023) | ✓ | RoBERTa, VGG-16 | Fusion method | ||
| Iparraguirre-Villanueva et al. (2023) | ✓ | Bag-of-words | LR, KNN, DT, RF, NB | ||
| *Proposed | ✓ | ✓ | ✓ | Fine-tuning of BERT, TF-IDF, word2vec | BERT classifier, RF, SVM |
Notes.
- IN
- Informative
- DM
- Damage
- IF
- Infrastructure
- H
- Human
- SVM
- Support Vector Machine
- RF
- Random Forest
- GB
- Gradient Boosting
- AB
- AdaBoost
- LSTM
- Long short-term Memory Network
- KNN
- K-nearest neighbor
- NB
- Naïve Bayes
- CNN
- Convolutional Neural Network
- GRU
- Gated Recurrent Unit
- AIDR
- Artificial Intelligence for Disaster Reponses
- DNN
- Deep Neural Network
- LR
- Logistic Regression
- DT
- Decision Tree