Search

In the task of human-machine dialogue for industrial operation and maintenance, in order to solve problems such as complex nested entities, missing words and typos in the data, this paper proposed an improved BERT joint algorithm GPGraphBERT, which used the correlation of intent and semantic slot recognition to improve dialogue performance. Firstly, after obtaining the hidden layer states from BERT, it constructed an adjacency matrix to convert them into a graph structure and embedded it into WRGAT to enhance the model's neighbor-awareness capabilities. Secondly, the algorithm improved the GlobalPointer mechanism incorporating RoPE, enabling the model to uniformly recognize both regular and nested entities. Finally, it designed a joint loss function for intent recognition and semantic slot recognition tasks, leveraging their correlation to improve prediction accuracy. During model training, it introduced dynamic masking to enhance the model's robustness and generalization capabilities. Experimental results showed F₁ scores of GP-GraphBERT algorithm achieved 87.5% and 86.4% for intent recognition and semantic slot recognition on the industrial operation and maintenance human-machine dialogue datasets, which were respectively improved by 9.2% and 3.0% compared to the original network, while also met the requirements for nested entity identification. The experiments fully verifies the performance of the algorithm in the joint recognition task. [ABSTRACT FROM AUTHOR]

A spoken language understanding system is a crucial component of a dialogue system whose task is to comprehend the user’s verbal expressions and perform the corresponding tasks accordingly. Contextual spoken language understanding (contextual SLU) is an extremely critical issue on this field as it helps the system to understand the user’s verbal expressions more accurately, thus improving the system’s performance and accuracy. The aim of this paper is to enhance the effectiveness of contextual SLU analysis. Context-based language unit systems are mainly concerned with effectively integrating dialog context information. Current approaches usually use the same contextual information to guide the slot filling of all tokens, which may introduce irrelevant information and lead to comprehension bias and ambiguity. To solve this problem, we apply the principle of graph contrastive learning based on the graph convolutional network to enhance the model’s ability to aggregate contextual information. Simultaneously, applying graph contrastive learning can enhance the model’s effectiveness by strengthening its intention. More precisely, graph convolutional networks can consider contextual information and automatically aggregate contextual information, allowing the model to no longer rely on traditionally designed heuristic aggregation functions. The contrastive learning module utilizes the principle of contrastive learning to achieve the effect of intention enhancement, which can learn deeper semantic information and contextual relationships, and improve the model's effectiveness in three key tasks: slot filling, dialogue action recognition, and intention detection. Experiments on a synthetic dialogue dataset show that our model achieves state-of-the-art performance and significantly outperforms other previous approaches (Slot F1 values + 1.03% on Sim-M, + 2.32% on Sim-R; Act F1 values + 0.26% on Sim-M, + 0.56% on Sim-R; Frame Acc values + 3.15% on Sim-M, + 1.62% on Sim-R). The code is available at: . [ABSTRACT FROM AUTHOR]

Understanding spoken language is crucial for conversational agents, with intent detection and slot filling being the primary tasks in natural language understanding (NLU). Enhancing the NLU tasks can lead to an accurate and efficient virtual assistant thereby reducing the need for human intervention and expanding their applicability in other domains. Traditionally, these tasks have been addressed individually, but recent studies have highlighted their interconnection, suggesting better results when solved together. Recent advances in natural language processing have shown that pretrained word embeddings can enhance text representation and improve the generalization capabilities of models. However, the challenge of poor generalization in joint learning models for intent detection and slot filling remains due to limited annotated datasets. Additionally, traditional models face difficulties in capturing both the semantic and syntactic nuances of language, which are vital for accurate intent detection and slot filling. This study proposes a hybridized text representation method using a multichannel convolutional neural network with three embedding channels: non-contextual embeddings for semantic information, part-of-speech (POS) tag embeddings for syntactic features, and contextual embeddings for deeper contextual understanding. Specifically, we utilized word2vec for non-contextual embeddings, one-hot vectors for POS tags, and bidirectional encoder representations from transformers (BERT) for contextual embeddings. These embeddings are processed through a convolutional layer and a shared bidirectional long short-term memory (BiLSTM) network, followed by two softmax functions for intent detection and slot filling. Experiments on the air travel information system (ATIS) and SNIPS datasets demonstrated that our model significantly outperformed the baseline models, achieving an intent accuracy of 97.90% and slot filling F1-score of 98.86% on the ATIS dataset, and an intent accuracy of 98.88% and slot filling F1-score of 97.07% on the SNIPS dataset. These results highlight the effectiveness of our proposed approach in advancing dialogue systems, and paving the way for more accurate and efficient natural language understanding in real-world applications. [ABSTRACT FROM AUTHOR]

Natural language understanding is a foundational task in medical dialogue systems. However, there are still two key problems to be solved: (1) Multiple meanings of a word lead to ambiguity of intent; (2) character errors make slot entity extraction difficult. To solve the above problems, this paper proposes a character-word information interaction framework (CWIIF) for natural language understanding in the Chinese medical dialogue domain. The CWIIF framework contains an intent information adapter to solve the problem of intent ambiguity caused by multiple meanings of words in the intent detection task and a slot label extractor to solve the problem of difficulty in yellowslot entity extraction due to character errors in the slot filling task. The proposed framework is validated on two publicly available datasets, the Intelligent Medical Consultation System (IMCS-21) and Chinese Artificial Intelligence Speakers (CAIS). Experimental results from both datasets demonstrate that the proposed framework outperforms other baseline methods in handling Chinese medical dialogues. Notably, on the IMCS-21 dataset, precision improved by 2.42%, recall by 3.01%, and the F1 score by 2.4%. [ABSTRACT FROM AUTHOR]

Natural Language Understanding (NLU) components are used in Dialog Systems (DS) to perform intent detection and entity extraction. In this work, we introduce a technique that exploits the inherent relationships between intents and entities to enhance the performance of NLU systems. The proposed method involves the utilization of a carefully crafted set of rules that formally express these relationships. By utilizing these rules, we effectively address inconsistencies within the NLU output, leading to improved accuracy and reliability. We implemented the proposed method using the Rasa framework as an NLU component and used our own conversational dataset AWPS to evaluate the improvement. Then, we validated the results in other three commonly used datasets: ATIS, SNIPS, and NLU-Benchmark. The experimental results show that the proposed method has a positive impact on the semantic accuracy metric, reaching an improvement of 12.6% in AWPS when training with a small amount of data. Furthermore, the practical application of the proposed method can easily be extended to other Task-Oriented Dialog Systems (T-ODS) to boost their performance and enhance user satisfaction. [ABSTRACT FROM AUTHOR]

Building task-oriented dialogue systems has become a topic of interest in the research community and industry. The task-oriented dialogue system is a closed-domain dialogue system that can perform specific tasks for users. The natural language understanding module of a task-oriented dialogue system is crucial because it is related to a task-oriented dialogue system that provides correctional services for users. The natural language understanding module of a task-oriented dialogue system performs two tasks: intent detection and slot filling. The intent detection task can be regarded as a text classification task; a classification model is trained to predict the intention of the user from the user's input information. The slot filling task can be regarded as a sequence analysis task; a sequence analysis model is trained to predict the details of the user's intention. In this paper, we proposed a novel model based on a transformer encoder for intent detection and slot filling. It follows the encoder-decoder structure, including a vanilla Transformer encoder, a bidirectional LSTM encoder, a linear classification decoder for intent detection, and a conditional random field decoder for slot filling. The experimental results on two public datasets show that our proposed model outperforms the existing methods based on the Transformer and can be combined with BERT to achieve better intent detection and slot filling results. [ABSTRACT FROM AUTHOR]

Intent detection and slot filling are two crucial tasks for spoken language understanding, and they are closely related. The accuracy of spoken language understanding depends strongly on the effectiveness of the interaction between intent and slot representations. However, previous studies have primarily focused on exploring the interaction of intent and slot representations within individual utterances while neglecting the relevance of different utterances. The paper proposes the CEA-Net, which utilizes co-interactive external attention as its core mechanism to effectively capture information from multiple utterances and perform information interaction between the two tasks. Experimental results demonstrate that the CEA-Net achieves competitive results on the ATIS and SNIPS benchmarks while reducing the number of parameters by about 44% compared with the previous best open-source approach. Furthermore, since our framework models the correlation of multiple utterances, it shows promising effectiveness and robustness even with limited training resources or datasets. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Shanghai University / Shanghai Daxue Xuebao is the property of Journal of Shanghai University (Natural Sciences) Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Intent detection and slot filling have been two of the most vital tasks to question the crop diseases and pests using knowledge graph. Among them, intent detection is to determine the user's intention or purpose from the text entered by the user in the crop pest field. Slot filling is to identify and extract specific information from the user's text, especially for the slot, for example, "what is the control of tomato bacterial wilt?". The task identifies the semantic slot "tomato bacterial wilt" (head entity) and the intention "control " (relation) to form a triple (tomato bacterial wilt, control, -). Where"-" represents the semantic slot and the intention to find the answer from the knowledge graph. A graph convolution encoder has been proposed to capture the spatiotemporal semantic features of conversation discourse in recent years, in order to learn the co-occurrence relationship between intention detection and slot filling. The slot decoding context has improved the accuracy of intention detection. In turn, the target of intention detection can be optimized to further improve the performance of slot filling. The best performance has been achieved in the experiments on public data sets. Sub word attention adapter (SAA) has also been used to effectively extract the context features from complex tags, and then retain the overall discourse information. In addition, an intention attention adapter (IAA) has been proposed to obtain the comprehensive sentence features, in order to predict the users’ intention. In experiments, the knowledge question and answer have been focused mainly on the pipeline in the existing agricultural research. The named entity recognition has been firstly used to identify the head entity in the question. The text classification can be then used to classify the question for the intention. However, such a pipeline can be vulnerable to error propagation that caused by word segmentation errors. It is very necessary to make full use of the correlation between intent detection and slot filling in the agricultural field. In this study, a joint model of intent detection and slot filling was proposed to fully utilize the semantic information using CNN-Transformer (CDPCT-IDSF). According to the semantic complexity of crop pest text, the CNN network and multi-layer Transformer were designed to emphasize the local useful information. The semantic loss was alleviated to introduce the alignment guarantee one-to-one relationship between input and output in the Transformer decoder. As such, the model was improved to identify the correct slot labels. In addition, a question-and-answer dataset was constructed for the agricultural pest knowledge, including 20 intention categories, 12 slot categories, and 11242 labeled samples. Comparative experiments show that the slot-filling F1 value of the CDPCT-IDSF model on the corpus was 94.36%, the intent detection accuracy was 92.99%, and the overall recognition accuracy was 87.23%, which was better than other comparison models. The improved model was also performed better for the intent detection and slot filling in crop diseases and pests. The finding can provide the theoretical support to knowledge question and answer research on crop diseases and pests. Moreover, the experiments were also carried out on two public datasets. CDPCT-IDSF was also suitable for English corpus scenarios. On ATIS datasets, the slot filling F1 value was 95.43%, intent detection accuracy was 97.95%, and overall recognition accuracy was 87.68%. On SNIPS datasets, the slot filling F1 value, intent detection accuracy and overall recognition accuracy of CDPCT-IDSF were 95.70%, 98.57% and 89.96%, respectively. At the same time, the better generalization and robustness were achieved in the rest corpus. [ABSTRACT FROM AUTHOR]

To address the lack of annotated data and insufficient performance in intent detection and slot filling in the domain of industrial equipment failure, a joint model based on BERT was proposed. BERT was utilized for text sequence encoding, while a bidirectional long short-term memory (Bi-LSTM) network was employed to capture the semantic relationships within the context. Max pooling and dense layers were used to extract key information, and a conditional random field (CRF) was incorporated to enhance the model's generalization capability. A question-and-answer corpus specifically tailored to the industrial domain of equipment failure was constructed, and a deployment framework for this domain was proposed. Experimental evaluations conducted on public datasets such as ATIS demonstrated that the proposed model outperforms baseline models by improving sentence-level accuracy, F1 score, and intent detection accuracy by 4. 4%, 2. 1%, and 0. 5% respectively. This research effectively enhances the performance of question-and-answer systems and provides a dataset and deployment framework for the field of industrial equipment failure, which lacks sufficient real-world data. [ABSTRACT FROM AUTHOR]

Understanding spoken language is crucial for conversational agents, with intent detection and slot filling being the primary tasks in natural language understanding (NLU). Enhancing the NLU tasks can lead to an accurate and efficient virtual assistant thereby reducing the need for human intervention and expanding their applicability in other domains. Traditionally, these tasks have been addressed individually, but recent studies have highlighted their interconnection, suggesting better results when solved together. Recent advances in natural language processing have shown that pretrained word embeddings can enhance text representation and improve the generalization capabilities of models. However, the challenge of poor generalization in joint learning models for intent detection and slot filling remains due to limited annotated datasets. Additionally, traditional models face difficulties in capturing both the semantic and syntactic nuances of language, which are vital for accurate intent detection and slot filling. This study proposes a hybridized text representation method using a multichannel convolutional neural network with three embedding channels: non-contextual embeddings for semantic information, part-of-speech (POS) tag embeddings for syntactic features, and contextual embeddings for deeper contextual understanding. Specifically, we utilized word2vec for non-contextual embeddings, one-hot vectors for POS tags, and bidirectional encoder representations from transformers (BERT) for contextual embeddings. These embeddings are processed through a convolutional layer and a shared bidirectional long short-term memory (BiLSTM) network, followed by two softmax functions for intent detection and slot filling. Experiments on the air travel information system (ATIS) and SNIPS datasets demonstrated that our model significantly outperformed the baseline models, achieving an intent accuracy of 97.90% and slot filling F1-score of 98.86% on the ATIS dataset, and an intent accuracy of 98.88% and slot filling F1-score of 97.07% on the SNIPS dataset. These results highlight the effectiveness of our proposed approach in advancing dialogue systems, and paving the way for more accurate and efficient natural language understanding in real-world applications.

Natural language understanding is a crucial aspect of task-oriented dialogue systems, encompassing intent detection (ID) and slot filling (SF). Conventional approaches for ID and SF solve the problems in a separate manners, while recent studies are now leaning toward joint modeling to tackle multi-intent detection and SF. Although the advancements in prompt learning offer a unified framework for ID and SF, current prompt-based methods fail to fully exploit the semantics of intent and slot labels. Additionally, the potential of using prompt learning to model the correlation between ID and SF in multi-intent scenarios remains unexplored. To address the issue, we propose a text-generative framework that unifies ID and SF. The prompt templates are constructed with label semantical descriptions. Moreover, we introduce an auxiliary task to explicitly capture the correlation between ID and SF. The experimental results on two benchmark datasets show that our method achieves an overall accuracy improvement of 0.4 – 1.5 % in a full-data scenario and 1.4 – 2.7 % in a few-shot setting compared with a prior method, establishing it as a new state-of-the-art approach. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Cross-lingual spoken language understanding (cross-lingual SLU), as a key component of task-oriented dialogue systems, is widely used in various industrial and real-world scenarios, such as multilingual customer support systems, cross-border communication platforms, and international language learning tools. However, obtaining large-scale and high-quality datasets for SLU is challenging due to the high cost of dialogue collection and manual annotation, particularly for minority languages. As a result, there is increasing interest in leveraging high-resource language data for cross-lingual transfer learning. Existing approaches for zero-shot cross-lingual SLU primarily focus on the relationship between the source language sentence and the single generated cross-lingual sentence, disregarding the shared information among multiple languages. This limitation weakens the robustness of multilingual word embedding representations and hampers the scalability of the model. In this paper, we propose the multilingual mixture attention interaction framework with adversarial training to alleviate the above problems. Specifically, we leverage the source language sentence to generate multiple multilingual hybrid sentences, in which words can adaptively capture unambiguous representations from the aligned multilingual words during the encoding phase, and adversarial training is introduced to enhance the scalability of the model. Then, we incorporate the symmetric kernel self-attention module with positional embedding to learn contextual information within a sentence, and employ the multi-relation graph convolutional networks to learn different granularity information between two highly correlated intent detection and slot filling tasks. Experimental results on the public dataset MultiATIS++ demonstrate that our proposed model achieves state-of-the-art performance, and comprehensive analysis validates the effectiveness of each component. [ABSTRACT FROM AUTHOR]

Information retrieval from databases is challenging for a non-SQL-domain expert. Some previous studies have provided solutions for translating the natural language to SQL instruction, aiming to access the information in the database directly. However, most solutions are in English Natural Language. In addition, the accuracies of the existing works still need to be improved. This work presents a mechanism called SV2-SQL, based on the pre-trained BERT. The proposed SV2-SQL mainly consists of multiple deep-learning models, including select-where slot filling model (SWSF-model), value extraction model (VE-model), and verification (V-model). The SWSF-model handles the classification tasks for those fields that appear in the “Select” and “Where” clauses, and the VE-model extracts the values for the “Where” clause from the input. The V-model sorts out the unwanted candidates from two previous models and leaves only the ones with the highest possibility. The proposed SV2-SQL also includes an algorithm for the inference process and allows the three models to be cooperative. Experimental results show that the proposed SV2-SQL outperforms the existing studies in terms of precision, accuracy, and recall. [ABSTRACT FROM AUTHOR]

Natural language understanding is a foundational task in medical dialogue systems. However, there are still two key problems to be solved: (1) Multiple meanings of a word lead to ambiguity of intent; (2) character errors make slot entity extraction difficult. To solve the above problems, this paper proposes a character-word information interaction framework (CWIIF) for natural language understanding in the Chinese medical dialogue domain. The CWIIF framework contains an intent information adapter to solve the problem of intent ambiguity caused by multiple meanings of words in the intent detection task and a slot label extractor to solve the problem of difficulty in yellowslot entity extraction due to character errors in the slot filling task. The proposed framework is validated on two publicly available datasets, the Intelligent Medical Consultation System (IMCS-21) and Chinese Artificial Intelligence Speakers (CAIS). Experimental results from both datasets demonstrate that the proposed framework outperforms other baseline methods in handling Chinese medical dialogues. Notably, on the IMCS-21 dataset, precision improved by 2.42%, recall by 3.01%, and the F1 score by 2.4%.

As the key equipment in power system, the running state of overhead transmission lines is affected by various complex and random factors, and the maintenance workload of the line is too heavy to overhaul regularly. Therefore, it is very necessary to build a refined condition assessment system to improve the diagnosis accuracy and maintenance efficiency of overhead transmission lines. Recent years, State Grid Corporation of China (SGCC) has recorded mass of monitoring reports of transmission lines. The natural language processing (NLP) with deep learning model provides an effective way to extract the key defect information from the monitoring reports. In this paper, the joint model of intention classification and slot filling (ICSF) based on bidirectional encoder representation from transformers (BERT) is introduced. To improve the precision of defect information extraction, two optimization models of BERT are presented. The results show that Robustly Optimized BERT Pre-Training Approach (RoBERTa) has achieved better effects on ICSF with the extraction accuracy of 92.22%. Then, the hierarchical weighted scoring method is introduced to score the status of overhead transmission line based on the results of ICSF-RoBERTa. And the assessment results of the overhead transmission line state and the corresponding maintenance strategies are provided according to the scores. Finally, the feasibility of the proposed method is validated by practical cases of line inspection reports.

As the core component of intelligent dialogue systems, spoken language understanding (SLU) usually includes two tasks: intent detection and slot filling. In real-world scenarios, users may express multiple intents in an utterance, and a token-level slot label can belong to multiple intents. Intent detection and slot filling tasks are closely related and instruct each other. In this paper, we propose the heterogeneous interaction graph framework with window mechanism for joint multi-intent detection and slot filling, which can adequately capture the rich semantic information of different granularity in heterogeneous information. We leverage different types of nodes and edges to construct the heterogeneous graph to realize the interaction between coarse-grained sentence-level intent information and fine-grained word-level slot information. And we utilize window mechanism to accommodate the temporal locality of the slot information. Experimental results on two datasets show that our model achieves the state-of-the-art performance. Comprehensive analysis empirically verifies the effectiveness of each component. [ABSTRACT FROM AUTHOR]

Intention detection and slot filling are two major subtasks in building a spoken language understanding (SLU) system. These two tasks are closely related to each other, and information from one will influence the other, establishing a bidirectional contributory relationship. Existing studies have typically modeled the two-way connection between these two tasks simultaneously in a unified framework. However, these studies have merely contributed to the research direction of fully using the correlations between feature information of the two tasks, without sufficient focusing on and utilizing native textual semantics. In this article, we propose a semantic guidance (SG) framework, enabling enhancing the understanding of textual semantics by dynamically gating the information from both tasks to acquire semantic features, ultimately leading to higher joint task accuracy. Experimental results on two widely used public datasets show that our model achieves state-of-the-art performance. [ABSTRACT FROM AUTHOR]

Slot filling, to extract entities for specific types of information (slot), is a vitally important modular of dialogue systems for automatic diagnosis. Doctor responses can be regarded as the weak supervision of patient queries. In this way, a large amount of weakly labeled data can be obtained from unlabeled diagnosis dialogue, alleviating the problem of costly and time-consuming data annotation. However, weakly labeled data suffers from extremely noisy samples. To alleviate the problem, we propose a simple and effective Co-Weak-Teaching method. The method trains two slot filling models simultaneously. These two models learn from two different weakly labeled data, ensuring learning from two aspects. Then, one model utilizes selected weakly labeled data generated by the other, iteratively. The model, obtained by the Co-Weak-Teaching on weakly labeled data, can be directly tested on testing data or sequentially fine-tuned on a small amount of human-annotated data. Experimental results on these two settings illustrate the effectiveness of the method with an increase of 8.03% and 14.74% in micro and macro f1 scores, respectively. [ABSTRACT FROM AUTHOR]

Intent detection and slot filling are two crucial subtasks of a text-based goal-oriented dialogue system. In a goal-oriented dialogue system, users interact with the system to complete a goal (or to fulfill their intent) and provide the necessary information (slot values) to achieve that goal. Therefore, a user's text input includes information about the user's intent and contains required slot values. Recently, joint models that simultaneously detect the intent and extract the slots are proposed to benefit from the interaction between the two tasks. The proposed methods are usually tested using benchmark data sets in English such as ATIS and SNIPS. Intent detection and slot filling problems are much less studied for the Turkish language mainly due to the lack of publicly available Turkish data sets. In this paper, we translate ATIS in English to Turkish and report intent detection and slot filling accuracies of several different joint models for the translated data set. We publicly share the Turkish ATIS data set to accelerate the research on the tasks. In our experiments, the best performance is obtained with the state-of-the-art bidirectional encoder representations from a transformers (BERT) based model. The BERT model is trained using a combination of intent detection and slot filling losses to jointly optimize a single model for both tasks. We achieved 96.54% intent detection accuracy and 91.56% slot filling F1 for the Turkish language. These accuracies significantly improve (7% absolute in slot filling F1) previously reported results for the same tasks in Turkish. On the other hand, we observe that the accuracy in Turkish is still slightly lower compared to the accuracy in English counterparts. This observation indicates that there is still room for improvement in the results for Turkish. [ABSTRACT FROM AUTHOR]

Intent detection and slot filling are two main tasks for building a spoken language understanding (SLU) system, the joint model of the two tasks is the research hotspot of the dialogue system. These two tasks are closely tied and the slots often highly depend on the intent. For the recent joint model: the fixed threshold is difficult to extract the positive votes in different domains. And the complex intent information guides the slot filling insufficiently. This paper proposes a fine-grained information integrated model for multiple intent detection and slot filling. Among this model, the intent information obtained by the intent decoder is concatenate with the encoding representation of each token to form an intent guided integrated encoding representation, so as to offer fine-grained intent information for the token-level slot prediction. At the same time, calculating the median of the maximum intent score and the minimum intent score to obtain the logic-adaptive threshold, and use it replace the fixed threshold. The logic-adaptive threshold can vary with the score distribution of different intent labels. Experimental results on two multi-label datasets verifies the performance of the proposed model. [ABSTRACT FROM AUTHOR]

• First investigation on machine reading comprehension framework for Chinese spoken language understanding. • Development of the TCM_CLS module with concise architecture for efficient multi-source information fusion. • Multi-task learning for joint slot filling and intent detection using adversarial training and focal loss. • Achievement of new SOTA performance compared with baseline methods. Spoken language understanding (SLU) plays a crucial role in the performance of dialogue systems. It usually includes slot filling and intent detection (SFID) tasks aiming at semantic parsing of utterances. At present, researchers focus mainly on English SLU tasks, while such investigations on Chinese utterances are not sufficient. In this paper, we first propose a reading comprehension based multi-source information fusion network, called ReCoMIF for Chinese SFID tasks by transforming the SLU task into a multi-turn question answering procedure comprising multiple choice for intent detection and span extraction for slot filling. Moreover, we present a TCM_CLS module with a concise architecture composed of TextCNN, MaxPooling, and feed forward network plus the [CLS] representation. Such three TCM_CLS modules are stacked in the proposed ReCoMIF network that can serve as sufficient integration of multi-source information originating from contexts, reading comprehension based queries, and hidden representations concerning intent and slot semantics. Finally, experimental results and ablation studies on three Chinese SLU datasets show that our proposed model can effectively fuse intent and slot information achieved by\ state-of-the-art performance compared with other baseline methods. [ABSTRACT FROM AUTHOR]

In task-based dialogue system, the key of the natural language understanding module is intent detection and slot filling. At this stage, Joint modeling of intention detection and slot filling tasks has become the mainstream and achieved good results. In order to investigate the correlation between intention detection and slot filling tasks, Joint model of intention detection and slot filling based on attention mechanism in three dimensions: one-way modeling from intention to slot, Unidirectional modeling from slot to intention and bidirectional modeling from intention to slotSeparately. And experiments were conducted using the Chinese dataset CAIS, and the results showed three evaluation results for time slot F1. The intention accuracy and overall accuracy of joint models for intention detection and filling gaps are usually higher than those of unidirectional models. [ABSTRACT FROM AUTHOR]

Human–robot interaction is becoming increasingly common to perform useful tasks in everyday life. From the human–machine communication perspective, achieving effective interaction in natural language is one challenge. To address it, natural language processing strategies have recently been used, commonly following a supervised machine learning framework. In this context, most approaches rely on the use of linguistic resources (e.g., taggers or embeddings), including training corpora. Unfortunately, such resources are scarce for some languages in specific domains, increasing the complexity of solution approaches. Motivated by these challenges, this paper explores deep learning methods for understanding natural language commands emitted to service robots that guide their movements in low-resource scenarios, defined by the use of Spanish and Nahuatl languages, for which linguistic resources are scarcely unavailable for this specific task. Particularly, we applied natural language understanding (NLU) techniques using deep neural networks and transformers-based models. As part of the research methodology, we introduced a labeled dataset of movement commands in the mentioned languages. The results show that models based on transformers work well to recognize commands (intent classification task) and their parameters (e.g., quantities and movement units) in Spanish, achieving a performance of 98.70% (accuracy) and 96.96% (F1) for the intent classification and slot-filling tasks, respectively). In Nahuatl, the best performance obtained was 93.5% (accuracy) and 88.57% (F1) in these tasks, respectively. In general, this study shows that robot movements can be guided in natural language through machine learning models using neural models and cross-lingual transfer strategies, even in low-resource scenarios.

With the integration of information technology and traditional industries,it has become a trend to use computer-controlled machines instead of humans to perform repetitive,boring and even dangerous tasks.In order to effectively interact with computers in natural language,human-computer interaction and dialogue systems based on multi-turn dialogue technology have become a research hotspot in the field of artificial intelligence and natural language processing.In the grid control system,the dispatcher needs to do a large number of query operations manually.To reduce the complexity of existing dispatching system and improve the speed of emergency handling of dispatchers,multi-turn dialogue technology can be applied to realize intelligent voice query of power grid data.This paper first describes the basic architecture of the task-oriented multi-turn dialogue system,including functions and related algorithms of its three modules:natural language understanding,dialogue management,and natural language generation.Next,in order to meet the demand of power grid companies for specific scenarios such as intelligent data queries,this paper designs and implements a multi-module task-oriented multi-turn dialogue system which consists of natural language understanding module,dialogue management module,natural language generation module and knowledge base as core mo-dules.The grid dispatcher can ask the system questions and get answers in the form of natural language.This process does not require keyboard or mouse operations,which greatly improves the rapidity and convenience of the grid information query.

As chatbots become more popular, multi-intent spoken language understanding (SLU) has received unprecedented attention. Multi-intent SLU, which primarily comprises the two subtasks of multiple intent detection (ID) and slot filling (SF), has the potential for widespread implementation. The two primary issues with the current approaches are as follows: (1) They cannot solve the problem of slot nesting; (2) The performance and inference rate of the model are not high enough. To address these issues, we suggest a multi-intent joint model based on global pointers to handle nested and non-nested slots. Firstly, we constructed a multi-dimensional type-slot label interaction network (MTLN) for subsequent intent decoding to enhance the implicit correlation between intents and slots, which allows for more adequate information about each other. Secondly, the global pointer network (GP) was introduced, which not only deals with nested and non-nested slots and slot incoherence but also has a faster inference rate and better performance than the baseline model. On two multi-intent datasets, the proposed model achieves state-of-the-art results on MixATIS with 1.6% improvement of intent Acc, 0.1% improvement of slot F1 values, 3.1% improvement of sentence Acc values, and 1.2%, 1.1% and 4.5% performance improvements on MixSNIPS, respectively. Meanwhile, the inference rate is also improved. [ABSTRACT FROM AUTHOR]

The comprehension of spoken language is a crucial aspect of dialogue systems, encompassing two fundamental tasks: intent classification and slot filling. Currently, the joint modeling approach for these two tasks has emerged as the dominant method in spoken language understanding modeling. However, the existing joint models have limitations in terms of their relevancy and utilization of contextual semantic features between the multiple tasks. To address these limitations, a joint model based on BERT and semantic fusion (JMBSF) is proposed. The model employs pre-trained BERT to extract semantic features and utilizes semantic fusion to associate and integrate this information. The results of experiments on two benchmark datasets, ATIS and Snips, in spoken language comprehension demonstrate that the proposed JMBSF model attains 98.80% and 99.71% intent classification accuracy, 98.25% and 97.24% slot-filling F1-score, and 93.40% and 93.57% sentence accuracy, respectively. These results reveal a significant improvement compared to other joint models. Furthermore, comprehensive ablation studies affirm the effectiveness of each component in the design of JMBSF. [ABSTRACT FROM AUTHOR]

Dialogue systems are becoming an ubiquitous presence in our everyday lives having a huge impact on business and society. Spoken language understanding (SLU) is the critical component of every goal-oriented dialogue system or any conversational system. The understanding of the user utterance is crucial for assisting the user in achieving their desired objectives. Future-generation systems need to be able to handle the multilinguality issue. Hence, the development of conversational agents becomes challenging as it needs to understand the different languages along with the semantic meaning of the given utterance. In this work, we propose a multilingual multitask approach to fuse the two primary SLU tasks, namely, intent detection and slot filling for three different languages. While intent detection deals with identifying user's goal or purpose, slot filling captures the appropriate user utterance information in the form of slots. As both of these tasks are highly correlated, we propose a multitask strategy to tackle these two tasks concurrently. We employ a transformer as a shared sentence encoder for the three languages, i.e., English, Hindi, and Bengali. Experimental results show that the proposed model achieves an improvement for all the languages for both the tasks of SLU. The multi-lingual multi-task (MLMT) framework shows an improvement of more than 2% in case of intent accuracy and 3% for slot F1 score in comparison to the single task models. Also, there is an increase of more than 1 point intent accuracy and 2 points slot F1 score in the MLMT model as opposed to the language specific frameworks. • We propose a multilingual multitask approach to fuse the two primary SLU tasks. • We propose a multitask strategy to tackle these two tasks concurrently. • We employ a transformer as a shared sentence encoder for three languages. [ABSTRACT FROM AUTHOR]

Manually annotating fine-grained slot-value labels for task-oriented dialogue (ToD) systems is an expensive and time-consuming endeavour. This motivates research into slot-filling methods that operate with limited amounts of labelled data. Moreover, the majority of current work on ToD is based solely on text as the input modality, neglecting the additional challenges of imperfect automatic speech recognition (ASR) when working with spoken language. In this work, we propose a Knowledge-Aware Audio-Grounded generative slot filling framework, termed KA2G, that focuses on few-shot and zero-shot slot filling for ToD with speech input. KA2G achieves robust and data-efficient slot filling for speech-based ToD by (1) framing it as a text generation task, (2) grounding text generation additionally in the audio modality, and (3) conditioning on available external knowledge (e.g. a predefined list of possible slot values). We show that combining both modalities within the KA2G framework improves the robustness against ASR errors. Further, the knowledge-aware slot-value generator in KA2G, implemented via a pointer generator mechanism, particularly benefits few-shot and zero-shot learning. Experiments, conducted on the standard speech-based single-turn SLURP dataset and a multi-turn dataset extracted from a commercial ToD system, display strong and consistent gains over prior work, especially in few-shot and zero-shot setups. • A knowledge-aware audio-grounded (KA2G) generative slot-filling framework is proposed • KA2G integrates knowledge with two tree-constrained pointer generator (TCPGen) • 4.6% and 11.2% SLU-F1 increases achieved for rare and unseen entities respectively. • KA2G achieved 20% joint goal accuracy (JGA) improvements on multi-turn dialogue. • The importance of the two TCPGen components were verified via comprehensive analyses [ABSTRACT FROM AUTHOR]

Intention detection and slot filling are two major subtasks in building a spoken language understanding (SLU) system. These two tasks are closely related to each other, and information from one will influence the other, establishing a bidirectional contributory relationship. Existing studies have typically modeled the two-way connection between these two tasks simultaneously in a unified framework. However, these studies have merely contributed to the research direction of fully using the correlations between feature information of the two tasks, without sufficient focusing on and utilizing native textual semantics. In this article, we propose a semantic guidance (SG) framework, enabling enhancing the understanding of textual semantics by dynamically gating the information from both tasks to acquire semantic features, ultimately leading to higher joint task accuracy. Experimental results on two widely used public datasets show that our model achieves state-of-the-art performance.

Cloud/edge computing and deep learning greatly improve performance of semantic understanding systems, where cloud/edge computing provides flexible, pervasive computation and storage capabilities to support variant applications, and deep learning models could comprehend text inputs by consuming computing and storage resource. Therefore, we propose to implement an intelligent online custom service system with power of both technologies. Essentially, task of semantic understanding consists of two subtasks, i.e., intent recognition and slot filling. To prevent error accumulation caused by modeling two subtasks independently, we propose to jointly model both subtasks in an end-to-end neural network. Specifically, the proposed method firstly extracts distinctive features with a dual structure to take full advantage of interactive and level information between two sub-tasks. Afterwards, we introduce attention scheme to enhance feature representation by involving sentence-level context information. With the support of cloud/edge computing infrastructure, we deploy the proposed network to work as an intelligent dialogue system for electrical customer service. During experiments, we test the proposed method and several comparative studies on public ATIS and our collected PSCF dataset. Experiment results prove the effectiveness of the proposed method by obtaining accurate and promising results. [ABSTRACT FROM AUTHOR]