Incremental Few-shot Text Classification with Multi-round New Classes: Formulation, Dataset and System

By Congyin Xia, Wenpeng Yin, Yihao Feng, Philip S. Yu. 2021

54 citations recorded by the source corpus.

Text classification is usually studied by labeling natural language texts with relevant categories from a predefined set. In the real world, new classes might keep challenging the existing system with limited labeled data. The system should be intelligent enough to recognize upcoming new classes with a few examples. In this work, we define a new task in the NLP domain, incremental few-shot text classification, where the system incrementally handles multiple rounds of new classes. For each round, there is a batch of new classes with a few labeled examples per class. Two major challenges exist in this new task: (i) For the learning process, the system should incrementally learn new classes round by round without re-training on the examples of preceding classes; (ii) For the performance, the system should perform well on new classes without much loss on preceding classes. In addition to formulating the new task, we also release two benchmark datasets in the incremental few-shot setting: intent classification and relation classification. Moreover, we propose two entailment approaches, ENTAILMENT and HYBRID, which show promise for solving this novel problem.

Subfields: Artificial Intelligence, Natural Language Processing, Meta & Continual Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Zixuan Ke
2. Hu Xu
3. Lei Shu
4. Bin Liu
5. Jun Zhu
6. Bing Liu
7. Hang Su
8. Martin Wistuba

In Natural Language Processing

1. Lei Shu
2. Hu Xu
3. Zixuan Ke
4. Xingxing Zhang
5. Caiming Xiong
6. Wenpeng Yin
7. Jun Zhu
8. Hang Su

In Meta & Continual Learning

1. Zixuan Ke
2. Bin Liu
3. Hu Xu
4. Lei Shu
5. Liyuan Wang
6. Xingxing Zhang
7. Hang Su
8. Haowei Lin

Citations in the corpus

• A Comprehensive Survey of Continual Learning: Theory, Method and Application (2023, methodology): “The applications of continual learning in NLP are also highly diverse, such as dialogue system [130], [131], [260], [260], [282], [298], [437], text classification [87], [172], [325], [460], sentence generation [246], [298], [499], relation learning [343], [357], [455], [465], neural machine transla”
• Continual Training of Language Models for Few-Shot Learning (2022, background): “Several researchers have studied this problem recently (An-toniou et al., 2020; Qin and Joty, 2021; Jin et al., 2021; Xia et al., 2021; Yin et al., 2022).”
• ConTinTin: Continual Learning from Task Instructions (2022, background): “Xia et al. (2021) proposed few-shot CCL in which multi-round of new text tags (e.g., intents or relations expressed in the input text) are encountered sequentially, and each new tag is only accompanied by a couple of examples.”
• Memory Efficient Continual Learning with Transformers (2022, background): “Recent work studied catastrophic forgetting [55, 10, 27, 35] and incremental learning [66] for NLP and CV [32, 15].”
• Prompt Augmented Generative Replay via Supervised Contrastive Learning for Lifelong Intent Detection (2022, methodology): “…Ge nerative R eplay” ( PAGeR ) model where, similar to existing approaches (de Masson d'Autume et al., 2019; Wang et al., 2020; Qian et al., 2021; Xia et al., 2021; Wang et al., 2019; Obamuyide and Vlachos, 2019; Han et al., 2020) we do not store real samples in replay memory but instead store…”
• Continual Few-shot Intent Detection (2022, methodology): “The most recent work (Xia et al., 2021) provides the first study on few-shot class-incremental learning for intent detection (FSCIL-ID).”
• Memory Efficient Continual Learning for Neural Text Classification (2022, background): “Recent work studied catastrophic forgetting (Sun et al., 2019; Chuang et al., 2020; Ke et al., 2020; Liu et al., 2020) and incremental learning (Xia et al., 2021) for NLP. Pa-sunuru et al. (2021) focus on the few-shot setting where only a few data points are available for each task.”