Neural Network With Hierarchical Attention Mechanism for Contextual Topic Dialogue Generation

Sun, Xiao; Ding, B.

doi:10.1109/access.2022.3140820

Cited by 7 publications

(4 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Mostly, it can separate the significant m features. Hence, it can be efficiently decrease the parameters of the Then, m is represented as the given Equation (28).…”

Section: Global Aspect Based Feature Extraction Modulementioning

confidence: 99%

Designing a Recommendation Model by Dilated Convolutional Neural Network with Bi-Directional Gated Recurrent Unit

Mastan,

Al-Nuaimy

2024

Preprint

View full text Add to dashboard Cite

In recent years, the recommendation system is one of the important tools of e-commerce, which provides suggestions to the user for some resources such as hotels, songs, books, movies, etc. The existing hotel recommendation method has faced many problems such as data sparsity, cold-start problems, scalability, etc. So, this paper proposes the hotel recommendation system use the hybrid of dilated convolutional neural network (CNN) and bi-directional gated recurrent unit (BiGRU) with an attention mechanism. The main aim of this research is to develop a more efficient, scalable, regularized, and more generalized recommendation system which can recommend the name of the hotels to the travellers based on their preferences by analysing the previous so far traveller’s comments together with the rating value to improve the forecast accuracy. The aspect based attention mechanisms are employed to evaluate the word, sentence, and semantic level similarity weight based vectors for mining useful information. The proposed approach has shown improved performance than existing approaches in terms of 99.46% Accuracy, 98.94% Precision, RE98.84% Recall, and 98.75% F1-score.

show abstract

“…Mostly, it can separate the significant m features. Hence, it can be efficiently decrease the parameters of the Then, m is represented as the given Equation (28).…”

Section: Global Aspect Based Feature Extraction Modulementioning

confidence: 99%

Designing a Recommendation Model by Dilated Convolutional Neural Network with Bi-Directional Gated Recurrent Unit

Mastan,

Al-Nuaimy

2024

Preprint

View full text Add to dashboard Cite

show abstract

“…As generated responses tend to be generic, several methods are proposed to improve the diversity of the generated responses, e.g. by incorporating topic information [23] or using latent variable models and gate mechanism [24]. Li et al [25] train an encoder-decoder model in a bidirectional method by adding a backward reasoning step to avoid generic and dull responses.…”

Section: Related Workmentioning

confidence: 99%

Open Domain Response Generation Guided by Retrieved Conversations

Chen

Zhang²,

Chen³

et al. 2023

IEEE Access

View full text Add to dashboard Cite

Open domain response generation is the task of creating a response given a user query in any topics/domain. Limited by context and reference information, responses generated by current systems are often "bland" or generic. In this paper, we combine a response generation model with a retrieval system that searches for relevant utterances and responses. The generation model has two main components: a keyword extraction module and a two-stage transformer. The keyword extraction module aims to extract two types of keywords in an unsupervised fashion from the retrieved results: (1) keywords in the query not found in the retrieved utterances (DiffKey), and (2) overlapping keywords among the retrieved responses (SimKey). Given these keywords, the two-stage transformer first decides where to insert the keywords in the response, and the second generates the full response given the location of the keywords. The keyword extraction module and the two-stage transformer are connected in a single network, and so our system is trained end-to-end. Experimental results on Cornell Movie-Dialog corpus, Douban and Weibo demonstrate that our model outperforms state-of-the-art systems in terms of ROUGE, relevance scores and human evaluation.

show abstract

“…In this study, the contextual sentence encoding result was used as the input of the BERT model, and the Encoder part of the Transformer [ 6 ] model as the model framework. The model structure is shown in Figure 3 .…”

Section: Coding Model For Multi-turn Dialoguementioning

confidence: 99%

“…The retrieval-based method mainly completes the reply matching of each dialogue turn, a discriminative model. The neural generation-based models mainly include Sequence-to-Sequence (Seq2Seq) Models [ 3 , 4 , 5 ], Dialogue Context [ 6 ], Response Diversity [ 7 , 8 ], Topic and Personality [ 9 , 10 ], Outside Knowledge Base. Dialogue context, Response Diversity, Topic or Personality, and other methods adopt multi-classification of contextual dialogues and then select and integrate the best alternative answers to return.…”

Section: Introductionmentioning

confidence: 99%

Human–Machine Multi-Turn Language Dialogue Interaction Based on Deep Learning

Yang

et al. 2022

Micromachines

View full text Add to dashboard Cite

During multi-turn dialogue, with the increase in dialogue turns, the difficulty of intention recognition and the generation of the following sentence reply become more and more difficult. This paper mainly optimizes the context information extraction ability of the Seq2Seq Encoder in multi-turn dialogue modeling. We fuse the historical dialogue information and the current input statement information in the encoder to capture the context dialogue information better. Therefore, we propose a BERT-based fusion encoder ProBERT-To-GUR (PBTG) and an enhanced ELMO model 3-ELMO-Attention-GRU (3EAG). The two models mainly enhance the contextual information extraction capability of multi-turn dialogue. To verify the effectiveness of the two proposed models, we demonstrate the effectiveness of our model by combining data based on the LCCC-large multi-turn dialogue dataset and the Naturalconv multi-turn dataset. The experimental comparison results show that, in the multi-turn dialogue experiments of the open domain and fixed topic, the two Seq2Seq coding models proposed are significantly improved compared with the current state-of-the-art models. For specified topic multi-turn dialogue, the 3EAG model has the average BLEU value reaches the optimal 32.4, which achieves the best language generation effect, and the BLEU value in the actual dialogue verification experiment also surpasses 31.8. for open-domain multi-turn dialogue. The average BLEU value of the PBTG model reaches 31.8, the optimal 31.8 achieves the best language generation effect, and the BLEU value in the actual dialogue verification experiment surpasses 31.2. So, the 3EAG model is more suitable for fixed-topic multi-turn dialogues for the two tasks. The PBTG model is more muscular in open-domain multi-turn dialogue tasks; therefore, our model is significant for promoting multi-turn dialogue research.

show abstract

Neural Network With Hierarchical Attention Mechanism for Contextual Topic Dialogue Generation

Cited by 7 publications

References 27 publications

Designing a Recommendation Model by Dilated Convolutional Neural Network with Bi-Directional Gated Recurrent Unit

Designing a Recommendation Model by Dilated Convolutional Neural Network with Bi-Directional Gated Recurrent Unit

Open Domain Response Generation Guided by Retrieved Conversations

Human–Machine Multi-Turn Language Dialogue Interaction Based on Deep Learning

Contact Info

Product

Resources

About