A Domain-Specific Intelligent Chatbot Using Rasa with Enhanced Natural Language Understanding

References

[1] Rasa Technologies. Rasa NLU — Structured Intent Recognition with LLM Flexibility. rasa.com/nlu[20]

[2] Rasa Technologies. Rasa Architecture Overview. legacy-docs-oss.rasa.com[8]

[3] Vidiniotis, A. (2025). Rasa in 2025: Empowering Conversational AI with Open-Source. LinkedIn Pulse[4]

[4] India AI. (2021). Rasa Chatbot Framework – NLU/Core. indiaai.gov.in[7]

[5] Bešić et al. (2024). Enhancing Conversational AI with the Rasa Framework: Intent Understanding and NLU. JOMS WSGE[10]

[6] Swathi, D.R.G. et al. (2025). Enhancing Chatbot Responses Through Context-Aware NLU. JATIT, Vol. 103, No. 19[6]

[7] Arxiv (2020). Context-Aware NLU with Selective Attention (MTL-CNLU-SAWC). arxiv.org/abs/2506.01781[12]

[8] Bunk, T. et al. (2020). DIET: Lightweight Language Understanding for Dialogue Systems. arXiv:2004.09936[18]

[9] Rasa Technologies. (2020). Introducing DIET: State-of-the-Art Architecture Outperforms BERT. rasa.com blog[9]

[10] IJISAE. (2023). An Integrated DIET-BO Model for Intent Classification and Entity Extraction. ijisae.org[11]

[11] Rasa Community. (2022). Open Source Natural Language Processing. rasa.community[5]

[12] Rasa Technologies. (2020). Unpacking the TED Policy in Rasa Open Source. rasa.com blog[15]

[13] Rasa Technologies. (2020). Dialogue Policies in Rasa 2.0. rasa.com blog[14]

[14] Zenodo. (2022). Healthcare Chatbot using RASA. zenodo.org[13]

[15] Krishna, T. (2024). Building and Deploying a Rasa Chatbot. LinkedIn Pulse[2]

[16] GJETA. (2025). Implementation of an AI-powered FAQ Chatbot Using the Rasa Framework. gjeta.com[3]

[17] Chatbot Conference. (2024). Building Chatbots using AI, NLP, NLU, NLG & LLMs. chatbotconference.com[1]

[18] Rasa Technologies. (2025). How LLM Chatbot Architecture Works. rasa.com blog[19]

[19] Paper Length: ~4,800 words | Domain: Artificial Intelligence / Natural Language Processing | Keywords: Domain-Specific Chatbot, Rasa Framework, DIET Classifier, NLU Pipeline, Dialogue Management, TEDPolicy, Intent Classification, Entity Recognition, Transformer Models

[20] https://www.chatbotconference.com/ai-nlp-nlu-nlg

[21] https://www.linkedin.com/pulse/building-deploying-rasa-chatbot-tadi-krishna-ehbfc

[22] http://gjeta.com/sites/default/files/GJETA-2025-0193.pdf

[23] https://www.linkedin.com/pulse/rasa-2025-empowering-conversational-ai-open-source-vidiniotis-7ugdc

[24] https://rasa.community/open-source-nlu-nlp/

[25] https://www.jatit.org/volumes/Vol103No19/12Vol103No19.pdf

[26] https://indiaai.gov.in/article/rasa-chatbot-framework-nlu-core

[27] https://legacy-docs-oss.rasa.com/docs/rasa/arch-overview/

[28] https://rasa.com/blog/introducing-dual-intent-and-entity-transformer-diet-state-of-the-art-performance-on-a-lightweight-architecture

[29] https://www.jomswsge.com/Enhancing-conversational-ai-with-the-Rasa-framework-intent-understanding-and-NLU,191223,0,2.html

[30] https://ijisae.org/index.php/IJISAE/article/view/3602

[31] https://arxiv.org/html/2506.01781

[32] https://zenodo.org/records/6395568

[33] https://rasa.com/blog/dialogue-policies-rasa-2

[34] https://rasa.com/blog/unpacking-the-ted-policy-in-rasa-open-source

[35] https://rasa.com/docs/reference/primitives/intents-and-entities/

[36] https://rasa.com/docs/rasa/rules/

[37] https://arxiv.org/pdf/2004.09936.pdf

[38] https://rasa.com/blog/llm-chatbot-architecture

[39] https://rasa.com/nlu

[40] https://github.com/RasaHQ/rasa-demo/blob/main/data/nlu/nlu.yml

[41] https://www.geeksforgeeks.org/machine-learning/chatbots-using-python-and-rasa/

[42] https://forum.rasa.com/t/intents-do-not-obtain-nlu-threhold-because-of-domain-specificity-chatbot/3346

[43] https://rasa.com

[44] https://www.jetir.org/papers/JETIR2309320.pdf

[45] https://forum.rasa.com/t/does-setting-entity-recognition-false-affects-the-performance-of-intent-classification-task-of-diet/30447

[46] https://e-journal.unair.ac.id/JISEBI/article/download/56303/32184

[47] https://www.youtube.com/watch?v=YxMzz6NF6Zw

[48] https://github.com/WeiNyn/DIETClassifier-pytorch

@article{1716279,
      author = {P Bharat Chandra, Chinmaya Gouda, Kishan Kumar Bhuyan, Samraju Keshab, Swati Kanta Mishra},
      title = {A Domain-Specific Intelligent Chatbot Using Rasa with Enhanced Natural Language Understanding},
      journal = {Iconic Research And Engineering Journals},
      year = {2026},
      volume = {9},
      number = {10},
      pages = {1131-1137},
      issn = {2456-8880},
      url = {https://www.irejournals.com/formatedpaper/1716279.pdf},
      abstract = {This paper presents the design, implementation, and evaluation of a domain-specific intelligent chatbot built using the Rasa open-source framework, augmented with enhanced Natural Language Understanding (NLU) techniques. Modern conversational agents face significant challenges in accurately interpreting domain-constrained user queries — from intent ambiguity to multi-entity extraction in complex utterances. This work leverages Rasa's Dual Intent and Entity Transformer (DIET) classifier [10] alongside the Transformer Embedding Dialogue (TED) Policy to achieve high-performance intent classification and contextual dialogue management. The proposed system was evaluated on a healthcare/customer support domain dataset, achieving an intent detection accuracy of 92.1% and an entity recognition F1-score of 0.88. Experimental results validate the superiority of the context-aware pipeline over single-turn baseline NLU models, demonstrating a 25% improvement in human-centric evaluation metrics. The system demonstrates that combining Rasa's modular NLU pipeline with transformer-based embeddings enables scalable, interpretable, and accurate domain-specific conversational AI.[6]},
      month = {April},
      doi = {https://doi.org/10.64388/IREV9I10-1716279}
  }