Incremental Intent Detection for Medical Domain with Contrast Replay Networks

Bai, Ge; He, Shizhu; Liu, Kang; Zhao, Jun

doi:10.18653/v1/2022.findings-acl.280

Cited by 4 publications

(5 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Based on this problem, Qin et al constructed the GL-GIN model based on the graph attention network [16] and used non-autoregressive methods to alleviate the problem of inconsistent slots. Bai et al [23] proposed a memory based method to incrementally learn emerging intentions in order to address the high computational cost of storing new data and intentions each time and retraining the entire data. Jiang et al [28] proposed a method of separation parsing, which divides a sentence into multiple clauses containing a single intention, performs loop parsing on each clause, and finally integrates the parsing results.…”

Section: Related Workmentioning

confidence: 99%

“…The first one is the problem of data sparsity. The data sources for multi-intent detection are scarce [25], the amount of data is insufficient, and the cost of annotating data is very high [2], [23], making it difficult to obtain annotated data. Additionally, the occurrence frequency of some intents or slots is relatively low, which leads to poor detection performance of certain intents or slots.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Conceptual Knowledge Enhanced Model for Multi-Intent Detection and Slot Filling

HE,

ZHAO,

DUAN

et al. 2024

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

In Natural Language Understanding, intent detection and slot filling have been widely used to understand user queries. However, current methods tend to rely on single words and sentences to understand complex semantic concepts, and can only consider local information within the sentence. Therefore, they usually cannot capture long-distance dependencies well and are prone to problems where complex intentions in sentences are difficult to recognize. In order to solve the problem of longdistance dependency of the model, this paper uses ConceptNet as an external knowledge source and introduces its extensive semantic information into the multi-intent detection and slot filling model. Specifically, for a certain sentence, based on confidence scores and semantic relationships, the most relevant conceptual knowledge is selected to equip the sentence, and a concept context map with rich information is constructed. Then, the multi-head graph attention mechanism is used to strengthen context correlation and improve the semantic understanding ability of the model. The experimental results indicate that the model has significantly improved performance compared to other models on the MixATIS and MixSNIPS multi-intent datasets.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Conceptual Knowledge Enhanced Model for Multi-Intent Detection and Slot Filling

HE,

ZHAO,

DUAN

et al. 2024

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

show abstract

“…In incremental learning, a model could learn from both old tasks and new tasks, so that it could be effective in both old tasks and new tasks. There are many researches which employ incremental learning in various NLP tasks, such as a few-shot classincremental learning method for NER [20], a contrast replay network for intent detection [21], an incremental few-shot text classification method for text classification [22], and an incremental meta self-training method for RE [23]. In this paper, incremental learning is employed by fine-tuning the pre-trained model with training samples in different tasks and experimental results show that our proposed model is robust to various noisy questions.…”

Section: Related Workmentioning

confidence: 99%

A More Robust Model to Answer Noisy Questions in KBQA

Wang

Xü

et al. 2023

IEEE Access

View full text Add to dashboard Cite

In practical applications, the raw input to a Knowledge Based Question Answering (KBQA) system may vary in forms, expressions, sources, etc. As a result, the actual input to the system may contain various errors caused by various noise in raw data and processes of transmission, transformation, translation, etc. As a result, it is significant to evaluate and enhance the robustness of a KBQA model to various noisy questions. In this paper, we generate 29 datasets of various noisy questions based on the original SimpleQuestions dataset to evaluate and enhance the robustness of a KBQA model, and propose a model which is more robust to various noisy questions. Compared with traditional methods, the main contribution in this paper is that we propose a method of generating datasets of different noisy questions to evaluate the robustness of a KBQA model, and propose a KBQA model which contains incremental learning and Mask Language Model (MLM) in the question answering process, so that our model is less affected by different kinds of noise in questions and achieves higher accuracies on datasets of different noisy questions, which shows its robustness. Experimental results show that our model achieves an average accuracy of 78.1% on these datasets and outperforms the baseline BERT-based model by an average margin of 5.0% with the similar training cost. In addition, further experiments show that our model is compatible with other pre-trained models such as ALBERT and ELECTRA.

show abstract

“…Bao et al (2020) build a chat-bot framework using user intents. Bai et al (2022) aim at incremental medical intent detection. Razzaq et al (2017); Amato et al (2017) develop an e-Health application using intent-context relationships.…”

Section: Related Workmentioning

confidence: 99%

Intent Identification and Entity Extraction for Healthcare Queries in Indic Languages

Mullick¹,

Mondal²,

Ray³

et al. 2023

Preprint

View full text Add to dashboard Cite

Scarcity of data and technological limitations for resource-poor languages in developing countries like India poses a threat to the development of sophisticated NLU systems for healthcare. To assess the current status of various state-of-the-art language models in healthcare, this paper studies the problem by initially proposing two different Healthcare datasets, Indian Healthcare Query Intent-WebMD and 1mg (IHQID-WebMD and IHQID-1mg) and one real world Indian hospital query data in English and multiple Indic languages (Hindi, Bengali, Tamil, Telugu, Marathi and Gujarati) which are annotated with the query intents as well as entities. Our aim is to detect query intents and extract corresponding entities. We perform extensive experiments on a set of models in various realistic settings and explore two scenarios based on the access to English data only (less costly) and access to target language data (more expensive). We analyze context specific practical relevancy through empirical analysis. The results, expressed in terms of overall F1 score show that our approach is practically useful to identify intents and entities.

show abstract

Incremental Intent Detection for Medical Domain with Contrast Replay Networks

Cited by 4 publications

References 20 publications

Conceptual Knowledge Enhanced Model for Multi-Intent Detection and Slot Filling

Conceptual Knowledge Enhanced Model for Multi-Intent Detection and Slot Filling

A More Robust Model to Answer Noisy Questions in KBQA

Intent Identification and Entity Extraction for Healthcare Queries in Indic Languages

Contact Info

Product

Resources

About