no code implementations • Findings (EMNLP) 2021 • Guy Feigenblat, Chulaka Gunasekara, Benjamin Sznajder, Sachindra Joshi, David Konopnicki, Ranit Aharonov
In most cases, at the end of the conversation, agents are asked to write a short summary emphasizing the problem and the proposed solution, usually for the benefit of other agents that may have to deal with the same customer or issue.
Extractive Summarization
Unsupervised Extractive Summarization
no code implementations • EMNLP 2020 • Nikhil Verma, Abhishek Sharma, Dhiraj Madan, Danish Contractor, Harshit Kumar, Sachindra Joshi
Neural Conversational QA tasks such as ShARC require systems to answer questions based on the contents of a given passage.
1 code implementation • Findings (EMNLP) 2021 • Chulaka Gunasekara, Guy Feigenblat, Benjamin Sznajder, Ranit Aharonov, Sachindra Joshi
Particularly, the results from human evaluations show that the summaries generated by our approach is preferred over 30% of the time over the summaries generated by general abstractive summarization models.
no code implementations • 12 Feb 2025 • Kushagra Bhushan, Yatin Nandwani, Dinesh Khandelwal, Sonam Gupta, Gaurav Pandey, Dinesh Raghu, Sachindra Joshi
While RAG enhances response relevance by incorporating retrieved domain knowledge in the context, retrieval errors can still lead to hallucinations and incorrect answers.
no code implementations • 7 Sep 2024 • Sonam Gupta, Yatin Nandwani, Asaf Yehudai, Mayank Mishra, Gaurav Pandey, Dinesh Raghu, Sachindra Joshi
The results show that standard SFT can lead to an average performance drop of up to $16. 7\%$ on multiple benchmarks, such as MMLU and TruthfulQA.
no code implementations • 13 Jun 2024 • G P Shrivatsa Bhargav, Sumit Neelam, Udit Sharma, Shajith Ikbal, Dheeraj Sreedhar, Hima Karanam, Sachindra Joshi, Pankaj Dhoolia, Dinesh Garg, Kyle Croutwater, Haode Qi, Eric Wayne, J William Murdock
The fine-tuning data is prepared carefully to cover a wide variety of slot-filling task scenarios that the model is expected to face across various domains.
1 code implementation • 7 Mar 2024 • Saeel Sandeep Nachane, Ojas Gramopadhye, Prateek Chanda, Ganesh Ramakrishnan, Kshitij Sharad Jadhav, Yatin Nandwani, Dinesh Raghu, Sachindra Joshi
In this paper, we propose a modified version of the MedQA-USMLE dataset, named MEDQA-OPEN, which contains open-ended medical questions without options to mimic clinical scenarios, along with clinician-approved reasoned answers.
no code implementations • 4 Feb 2024 • Gaurav Pandey, Yatin Nandwani, Tahira Naseem, Mayank Mishra, Guangxuan Xu, Dinesh Raghu, Sachindra Joshi, Asim Munawar, Ramón Fernandez Astudillo
Distribution matching methods for language model alignment such as Generation with Distributional Control (GDC) and Distributional Policy Gradient (DPG) have not received the same level of attention in reinforcement learning from human feedback (RLHF) as contrastive methods such as Sequence Likelihood Calibration (SLiC), Direct Preference Optimization (DPO) and its variants.
no code implementations • 15 Nov 2023 • Ankita Gupta, Chulaka Gunasekara, Hui Wan, Jatin Ganhotra, Sachindra Joshi, Marina Danilevsky
We find that both fine-tuned and instruction-tuned models are affected by input variations, with the latter being more susceptible, particularly to dialogue-level perturbations.
no code implementations • 9 Sep 2023 • Biplav Srivastava, Kausik Lakkaraju, Tarmo Koppel, Vignesh Narayanan, Ashish Kundu, Sachindra Joshi
Chatbots, the common moniker for collaborative assistants, are Artificial Intelligence (AI) software that enables people to naturally interact with them to get tasks done.
1 code implementation • 20 May 2023 • Yatin Nandwani, Vineet Kumar, Dinesh Raghu, Sachindra Joshi, Luis A. Lastras
PMI quantifies the extent to which the document influences the generated response -- with a higher PMI indicating a more faithful response.
1 code implementation • NAACL 2022 • Pengshan Cai, Hui Wan, Fei Liu, Mo Yu, Hong Yu, Sachindra Joshi
We propose novel AI-empowered chat bots for learning as conversation where a user does not read a passage but gains information and knowledge through conversation with a teacher bot.
1 code implementation • NAACL (ACL) 2022 • Hui Wan, Siva Sankalp Patel, J. William Murdock, Saloni Potdar, Sachindra Joshi
Dialogue systems can benefit from being able to search through a corpus of text to find information relevant to user requests, especially when encountering a request for which no manually curated response is available.
no code implementations • 6 Apr 2022 • Gaurav Pandey, Danish Contractor, Sachindra Joshi
In this paper, we combine the best of both worlds by proposing a scalable model that can learn complex relationships between context-response pairs.
no code implementations • SIGDIAL (ACL) 2022 • Qingyang Wu, Song Feng, Derek Chen, Sachindra Joshi, Luis A. Lastras, Zhou Yu
Collecting data for training dialog systems can be extremely expensive due to the involvement of human participants and need for extensive annotation.
1 code implementation • 23 Nov 2021 • Guy Feigenblat, Chulaka Gunasekara, Benjamin Sznajder, Sachindra Joshi, David Konopnicki, Ranit Aharonov
In most cases, at the end of the conversation, agents are asked to write a short summary emphasizing the problem and the proposed solution, usually for the benefit of other agents that may have to deal with the same customer or issue.
Extractive Summarization
Unsupervised Extractive Summarization
1 code implementation • EMNLP 2021 • Song Feng, Siva Sankalp Patel, Hui Wan, Sachindra Joshi
We propose MultiDoc2Dial, a new task and dataset on modeling goal-oriented dialogues grounded in multiple documents.
no code implementations • Findings (ACL) 2021 • Dinesh Raghu, Atishya Jain, Mausam, Sachindra Joshi
In this paper, we propose a novel filtering technique that consists of (1) a pairwise similarity based filter that identifies relevant information by respecting the n-ary structure in a KB record.
1 code implementation • EMNLP 2021 • Dinesh Raghu, Shantanu Agarwal, Sachindra Joshi, Mausam
We propose a novel problem within end-to-end learning of task-oriented dialogs (TOD), in which the dialog system mimics a troubleshooting agent who helps a user by diagnosing their problem (e. g., car not starting).
no code implementations • 18 Aug 2021 • Jatin Ganhotra, Samuel Thomas, Hong-Kwang J. Kuo, Sachindra Joshi, George Saon, Zoltán Tüske, Brian Kingsbury
End-to-end spoken language understanding (SLU) systems that process human-human or human-computer interactions are often context independent and process each turn of a conversation independently.
no code implementations • Findings (ACL) 2021 • Chulaka Gunasekara, Guy Feigenblat, Benjamin Sznajder, Sachindra Joshi, David Konopnicki
Many conversation datasets have been constructed in the recent years using crowdsourcing.
no code implementations • NAACL 2021 • Hui Wan, Song Feng, Chulaka Gunasekara, Siva Sankalp Patel, Sachindra Joshi, Luis Lastras
Machine reading comprehension is a challenging task especially for querying documents with deep and interconnected contexts.
1 code implementation • NAACL 2021 • Hanjie Chen, Song Feng, Jatin Ganhotra, Hui Wan, Chulaka Gunasekara, Sachindra Joshi, Yangfeng Ji
Most existing methods generate post-hoc explanations for neural network models by identifying individual feature attributions or detecting interactions between adjacent features.
no code implementations • 24 Jan 2021 • Jatin Ganhotra, Sachindra Joshi
However, there has been little to no research on the impact of this limitation with respect to dialog tasks.
2 code implementations • EMNLP 2020 • Song Feng, Hui Wan, Chulaka Gunasekara, Siva Sankalp Patel, Sachindra Joshi, Luis A. Lastras
We introduce doc2dial, a new dataset of goal-oriented dialogues that are grounded in the associated documents.
no code implementations • Findings (EMNLP) 2021 • Biswesh Mohapatra, Gaurav Pandey, Danish Contractor, Sachindra Joshi
Popular dialog datasets such as MultiWOZ are created by providing crowd workers an instruction, expressed in natural language, that describes the task to be accomplished.
1 code implementation • Findings of the Association for Computational Linguistics 2020 • Jatin Ganhotra, Robert Moore, Sachindra Joshi, Kahini Wadhawan
Existing benchmarks used to evaluate the performance of end-to-end neural dialog systems lack a key component: natural variation present in human conversations.
1 code implementation • EMNLP 2020 • Jatin Ganhotra, Haggai Roitman, Doron Cohen, Nathaniel Mills, Chulaka Gunasekara, Yosi Mass, Sachindra Joshi, Luis Lastras, David Konopnicki
A frequent pattern in customer care conversations is the agents responding with appropriate webpage URLs that address users' needs.
no code implementations • EMNLP 2020 • Kshitij Fadnis, Nathaniel Mills, Jatin Ganhotra, Haggai Roitman, Gaurav Pandey, Doron Cohen, Yosi Mass, Shai Erera, Chulaka Gunasekara, Danish Contractor, Siva Patel, Q. Vera Liao, Sachindra Joshi, Luis Lastras, David Konopnicki
Customer support agents play a crucial role as an interface between an organization and its end-users.
no code implementations • 11 Feb 2020 • Gaurav Pandey, Dinesh Raghu, Sachindra Joshi
The proposed model, referred to as Mask \& Focus maps the input context to a sequence of concepts which are then used to generate the response concepts.
no code implementations • IJCNLP 2019 • Harshit Kumar, Arvind Agarwal, Sachindra Joshi
This paper proposes an end-to-end multi-task model for conversation modeling, which is optimized for two tasks, dialogue act prediction and response selection, with the latter being the task of interest.
2 code implementations • 9 Sep 2019 • Nikhil Verma, Abhishek Sharma, Dhiraj Madan, Danish Contractor, Harshit Kumar, Sachindra Joshi
On studying recent state-of-the-art models on the ShARCQA task, we found indications that the models learn spurious clues/patterns in the dataset.
no code implementations • 2 Nov 2018 • Dhiraj Madan, Dinesh Raghu, Gaurav Pandey, Sachindra Joshi
However these states need to be handcrafted and annotated in the data.
1 code implementation • NAACL 2019 • Revanth Reddy, Danish Contractor, Dinesh Raghu, Sachindra Joshi
Instead of using triples to store KB results, we introduce a novel multi-level memory architecture consisting of cells for each query and their corresponding results.
no code implementations • COLING 2018 • Harshit Kumar, Arvind Agarwal, Sachindra Joshi
The utility of additional semantic information for the task of next utterance selection in an automated dialogue system is the focus of study in this paper.
no code implementations • 3 Jul 2018 • Ayushi Dalmia, Sachindra Joshi, Raghavendra Singh, Vikas Raykar
We pose the problem of predicting complementary fashion items as a sequence to sequence problem where the input is the selected set of fashion items and the output is a complementary fashion item based on the style information learned by the model.
no code implementations • ACL 2018 • P, Gaurav ey, Danish Contractor, Vineet Kumar, Sachindra Joshi
In this paper we present the Exemplar Encoder-Decoder network (EED), a novel conversation model that learns to utilize \textit{similar} examples from training data to generate responses.
no code implementations • IJCNLP 2017 • Dhiraj Madan, Sachindra Joshi
We propose a novel SimCluster algorithm that extends standard K-means algorithm to simultaneously cluster user utterances and agent utterances by taking their adjacency information into account.
3 code implementations • 13 Sep 2017 • Harshit Kumar, Arvind Agarwal, Riddhiman Dasgupta, Sachindra Joshi, Arun Kumar
Dialogue Act recognition associate dialogue acts (i. e., semantic labels) to utterances in a conversation.
no code implementations • EACL 2017 • Sathish Reddy, Dinesh Raghu, Mitesh M. Khapra, Sachindra Joshi
To generate such QA pairs, we first extract a set of keywords from entities and relationships expressed in a triple stored in the knowledge graph.
no code implementations • COLING 2016 • Vineet Kumar, Sachindra Joshi
The system thus needs to take into account the conversation context to process the question.
no code implementations • LREC 2014 • Subhabrata Mukherjee, Sachindra Joshi
Furthermore, we also show the effectiveness of our approach in capturing thwarting in reviews, achieving an accuracy improvement of 11. 53{\%} over the SVM baseline.