Conversational AI is a subtype of Human Computer Interaction that has gained wide adoption. These systems are typically powered by Large Language Models (LLMs) that use Retrieval Augmented Generation (RAG) to infuse external knowledge, which is effective against issues like hallucination. However, automatically evaluating retrieval augmented conversations with minimal human effort remains challenging, particularly in online settings. We address this challenge by proposing a lexical metric, and a novel method for combining it with other metrics, including semantic models. Our approach involves: (1) Conversational Information Utility (CIU), a new automated metric inspired by prior user studies on web search evaluation, to compute information overlap between conversation context and grounded information in an unsupervised, purely lexical way; and (2) a generalized reward model through Mixture-of-Experts (MoE-CIU) that dynamically ensembles CIU with other metrics, including learned ones, into a single reward. Evaluation against human ratings on two public datasets (Topical Chat and Persona Chat) shows that CIU improves correlation against human judgments by 2.0% and 0.9% respectively compared to the second best metric. When MoE is applied to combine lexical and learned semantic metrics, correlations further improve by 9.9% and 5.0%, suggesting that unified reward models are a promising approach.

Conversational Task Assistants (CTAs) guide users in performing a multitude of activities, such as making recipes. However, ensuring that interactions remain engaging, interesting, and enjoyable for CTA users is not trivial, especially for time-consuming or challenging tasks. Grounded in psychological theories of human interest, we propose to engage users with contextual and interesting statements or facts during interactions with a multi-modal CTA, to reduce fatigue and task abandonment before a task is complete. To operationalize this idea, we train a high-performing classifier (82% F1-score) to automatically identify relevant and interesting facts for users. We use it to create an annotated dataset of task-specific interesting facts for the domain of cooking. Finally, we design and validate a dialogue policy to incorporate the identified relevant and interesting facts into a conversation, to improve user engagement and task completion. Live testing on a leading multi-modal voice assistant shows that 66% of the presented facts were received positively, leading to a 40% gain in the user satisfaction rating, and a 37% increase in conversation length. These findings emphasize that strategically incorporating interesting facts into the CTA experience can promote real-world user participation for guided task interactions.

Continued improvement of conversational assistants in knowledge-rich domains like E-Commerce requires large volumes of realistic high-quality conversation data to power increasingly sophisticated large language model chatbots, dialogue managers, response rankers, and recommenders. The problem is exacerbated for multi-modal interactions in realistic conversational product search and recommendation. Here, an artificial sales agent must interact intelligently with a customer using both textual and visual information and incorporate results from external search systems, such as a product catalog. Yet, it remains an open question how to best crowd-source large-scale, naturalistic multi-modal dialogue and action data, required to train such an artificial agent. We describe our crowd-sourced task where one worker (the Buyer) plays the role of the customer, and another (the Seller) plays the role of the sales agent. We identify subtle interactions between one worker’s environment and their partner’s behavior mediated by workers’ word choice. We find that limiting information presented to the Buyer, both in their backstory and by the Seller, improves conversation quality. We also show how conversations are improved through minimal automated Seller “coaching”. While typed and spoken messages are slightly different, the differences are not as large as frequently assumed. We plan to release our platform code and the resulting dialogues to advance research on conversational search agents.

Large Language Models (LLMs) operating in 0-shot or few-shot settings achieve competitive results in Text Classification tasks. In-Context Learning (ICL) typically achieves better accuracy than the 0-shot setting, but it pays in terms of efficiency, due to the longer input prompt. In this paper, we propose a strategy to make LLMs as efficient as 0-shot text classifiers, while getting comparable or better accuracy than ICL. Our solution targets the low resource setting, i.e., when only 4 examples per class are available. Using a single LLM and few-shot real data we perform a sequence of generation, filtering and Parameter-Efficient Fine-Tuning steps to create a robust and efficient classifier. Experimental results show that our approach leads to competitive results on multiple text classification datasets.

We present the findings of SemEval-2023 Task 2 on Fine-grained Multilingual Named Entity Recognition (MultiCoNER 2). Divided into 13 tracks, the task focused on methods to identify complex fine-grained named entities (like WRITTENWORK, VEHICLE, MUSICALGRP) across 12 languages, in both monolingual and multilingual scenarios, as well as noisy settings. The task used the MultiCoNER V2 dataset, composed of 2.2 million instances in Bangla, Chinese, English, Farsi, French, German, Hindi, Italian., Portuguese, Spanish, Swedish, and Ukrainian. MultiCoNER 2 was one of the most popular tasks of SemEval-2023. It attracted 842 submissions from 47 teams, and 34 teams submitted system papers. Results showed that complex entity types such as media titles and product names were the most challenging. Methods fusing external knowledge into transformer models achieved the best performance, and the largest gains were on the Creative Work and Group classes, which are still challenging even with external knowledge. Some fine-grained classes proved to be more challenging than others, such as SCIENTIST, ARTWORK, and PRIVATECORP. We also observed that noisy data has a significant impact on model performance, with an average drop of 10% on the noisy subset. The task highlights the need for future research on improving NER robustness on noisy data containing complex entities.

Methods to generate text from structured data have advanced significantly in recent years, primarily due to fine-tuning of pre-trained language models on large datasets. However, such models can fail to produce output faithful to the input data, particularly on out-of-domain data. Sufficient annotated data is often not available for specific domains, leading us to seek an unsupervised approach to improve the faithfulness of output text. Since the problem is fundamentally one of consistency between the representations of the structured data and text, we evaluate the effectiveness of cycle training in this work. Cycle training uses two models which are inverses of each other: one that generates text from structured data, and one which generates the structured data from natural language text. We show that cycle training, when initialized with a small amount of supervised data (100 samples in our case), achieves nearly the same performance as fully supervised approaches for the data-to-text generation task on the WebNLG, E2E, WTQ, and WSQL datasets. We perform extensive empirical analysis with automated evaluation metrics and a newly designed human evaluation schema to reveal different cycle training strategies’ effectiveness of reducing various types of generation errors. Our code is publicly available at https://github.com/Edillower/CycleNLG.

Spoken Question Answering (QA) is a key feature of voice assistants, usually backed by multiple QA systems. Users ask questions via spontaneous speech that can contain disfluencies, errors, and informal syntax or phrasing. This is a major challenge in QA, causing unanswered questions or irrelevant answers, leading to bad user experiences. We analyze failed QA requests to identify core challenges: lexical gaps, proposition types, complex syntactic structure, and high specificity. We propose a Semantic Question Reformulation (SURF) model offering three linguistically-grounded operations (repair, syntactic reshaping, generalization) to rewrite questions to facilitate answering. Offline evaluation on 1M unanswered questions from a leading voice assistant shows that SURF significantly improves answer rates: up to 24% of previously unanswered questions obtain relevant answers (75%). Live deployment shows positive impact for millions of customers with unanswered questions; explicit relevance feedback shows high user satisfaction.

Frequently Asked Question (FAQ) retrieval aims at retrieving question-answer pairs for a given a user query. Integrating FAQ retrieval with product search can not only empower users to make more informed purchase decisions, but also enhance user retention through efficient post-purchase support. Providing FAQ content without disrupting user’s shopping experience poses challenges on deciding when and how to show FAQ results. Our proposed intent-aware FAQ retrieval consists of (1) an intent classifier that predicts whether the query is looking for an FAQ; (2) a reformulation model that rewrites query into a natural question. Offline evaluation demonstrates that our approach improves 12% in Hit@1 on retrieving ground-truth FAQs, while reducing latency by 95% compared to baseline systems. These improvements are further validated by real user feedback, where more than 99% of users consider FAQs displayed on top of product search results is helpful. Overall, our findings show promising directions for integrating FAQ retrieval into product search at scale.

The adoption of voice assistants like Alexa or Siri has grown rapidly, allowing users to instantly access information via voice search. Query suggestion is a standard feature of screen-based search experiences, allowing users to explore additional topics. However, this is not trivial to implement in voice-based settings. To enable this, we tackle the novel task of suggesting questions with compact and natural voice hints to allow users to ask follow-up questions. We define the task, ground it in syntactic theory and outline linguistic desiderata for spoken hints. We propose baselines and an approach using sequence-to-sequence Transformers to generate spoken hints from a list of questions. Using a new dataset of 6681 input questions and human written hints, we evaluated the models with automatic metrics and human evaluation. Results show that a naive approach of concatenating suggested questions creates poor voice hints. Our approach, which applies a linguistically-motivated pretraining task was strongly preferred by humans for producing the most natural hints.

We present MULTICONER V2, a dataset for fine-grained Named Entity Recognition covering 33 entity classes across 12 languages, in both monolingual and multilingual settings. This dataset aims to tackle the following practical challenges in NER: (i) effective handling of fine-grained classes that include complex entities like movie titles, and (ii) performance degradation due to noise generated from typing mistakes or OCR errors. The dataset is compiled from open resources like Wikipedia and Wikidata, and is publicly available. Evaluation based on the XLM-RoBERTa baseline highlights the unique challenges posed by MULTICONER V2: (i) the fine-grained taxonomy is challenging, where the scores are low with macro-F1=0.63 (across all languages), and (ii) the corruption strategy significantly impairs performance, with entity corruption resulting in 9% lower performance relative to non-entity corruptions across all languages. This highlights the greater impact of entity noise in contrast to context noise.

E-commerce product catalogs contain billions of items. Most products have lengthy titles, as sellers pack them with product attributes to improve retrieval, and highlight key product aspects. This results in a gap between such unnatural products titles, and how customers refer to them. It also limits how e-commerce stores can use these seller-provided titles for recommendation, QA, or review summarization. Inspired by recent work on instruction-tuned LLMs, we present InstructPTS, a controllable approach for the task of Product Title Summarization (PTS). Trained using a novel instruction fine-tuning strategy, our approach is able to summarize product titles according to various criteria (e.g. number of words in a summary, inclusion of specific phrases, etc.). Extensive evaluation on a real-world e-commerce catalog shows that compared to simple fine-tuning of LLMs, our proposed approach can generate more accurate product name summaries, with an improvement of over 14 and 8 BLEU and ROUGE points, respectively.

End-to-end neural models for conversational AI often assume that a response can be generated by considering only the knowledge acquired by the model during training. Document-oriented conversational models make a similar assumption by conditioning the input on the document and assuming that any other knowledge is captured in the model’s weights. However, a conversation may refer to external knowledge sources. In this work, we present EKo-Doc, an architecture for document-oriented conversations with access to external knowledge: we assume that a conversation is centered around a topic document and that external knowledge is needed to produce responses. EKo-Doc includes a dense passage retriever, a re-ranker, and a response generation model. We train the model end-to-end by using silver labels for the retrieval and re-ranking components that we automatically acquire from the attention signals of the response generation model. We demonstrate with automatic and human evaluations that incorporating external knowledge improves response generation in document-oriented conversations. Our architecture achieves new state-of-the-art results on the Wizard of Wikipedia dataset, outperforming a competitive baseline by 10.3% in Recall@1 and 7.4% in ROUGE-L.

Named entity recognition (NER) in a real-world setting remains challenging and is impacted by factors like text genre, corpus quality, and data availability. NER models trained on CoNLL do not transfer well to other domains, even within the same language. This is especially the case for multi-lingual models when applied to low-resource languages, and is mainly due to missing entity information. We propose an approach that with limited effort and data, addresses the NER knowledge gap across languages and domains. Our novel approach uses a token-level gating layer to augment pre-trained multilingual transformers with gazetteers containing named entities (NE) from a target language or domain. This approach provides the flexibility to jointly integrate both textual and gazetteer information dynamically: entity knowledge from gazetteers is used only when a token’s textual representation is insufficient for the NER task. Evaluation on several languages and domains demonstrates: (i) a high mismatch of reported NER performance on CoNLL vs. domain specific datasets, (ii) gazetteers significantly improve NER performance across languages and domains, and (iii) gazetteers can be flexibly incorporated to guide knowledge transfer. On cross-lingual transfer we achieve an improvement over the baseline with F1=+17.6%, and with F1=+21.3% for cross-domain transfer.

Users expect their queries to be answered by search systems, regardless of the query’s surface form, which include keyword queries and natural questions. Natural Language Understanding (NLU) components of Search and QA systems may fail to correctly interpret semantically equivalent inputs if this deviates from how the system was trained, leading to suboptimal understanding capabilities. We propose the keyword-question rewriting task to improve query understanding capabilities of NLU systems for all surface forms. To achieve this, we present CycleKQR, an unsupervised approach, enabling effective rewriting between keyword and question queries using non-parallel data.Empirically we show the impact on QA performance of unfamiliar query forms for open domain and Knowledge Base QA systems (trained on either keywords or natural language questions). We demonstrate how CycleKQR significantly improves QA performance by rewriting queries into the appropriate form, while at the same time retaining the original semantic meaning of input queries, allowing CycleKQR to improve performance by up to 3% over supervised baselines. Finally, we release a datasetof 66k keyword-question pairs.

Conversational Question Answering (CQA) aims to answer questions contained within dialogues, which are not easily interpretable without context. Developing a model to rewrite conversational questions into self-contained ones is an emerging solution in industry settings as it allows using existing single-turn QA systems to avoid training a CQA model from scratch. Previous work trains rewriting models using human rewrites as supervision. However, such objectives are disconnected with QA models and therefore more human-like rewrites do not guarantee better QA performance. In this paper we propose using QA feedback to supervise the rewriting model with reinforcement learning. Experiments show that our approach can effectively improve QA performance over baselines for both extractive and retrieval QA. Furthermore, human evaluation shows that our method can generate more accurate and detailed rewrites when compared to human annotations.

We describe an application of Knowledge Distillation used to distill and deploy multilingual Transformer models for voice assistants, enabling text classification for customers globally. Transformers have set new state-of-the-art results for tasks like intent classification, and multilingual models exploit cross-lingual transfer to allow serving requests across 100+ languages. However, their prohibitive inference time makes them impractical to deploy in real-world scenarios with low latency requirements, such as is the case of voice assistants. We address the problem of cross-architecture distillation of multilingual Transformers to simpler models, while maintaining multilinguality without performance degradation. Training multilingual student models has received little attention, and is our main focus. We show that a teacher-student framework, where the teacher’s unscaled activations (logits) on unlabelled data are used to supervise student model training, enables distillation of Transformers into efficient multilingual CNN models. Our student model achieves equivalent performance as the teacher, and outperforms a similar model trained on the labelled data used to train the teacher model. This approach has enabled us to accurately serve global customer requests at speed (18x improvement), scale, and low cost.

Factual and logical errors made by Natural Language Generation (NLG) systems limit their applicability in many settings. We study this problem in a conversational search and recommendation setting, and observe that we can often make two simplifying assumptions in this domain: (i) there exists a body of structured knowledge we can use for verifying factuality of generated text; and (ii) the text to be factually assessed typically has a well-defined structure and style. Grounded in these assumptions, we propose a fast, unsupervised and explainable technique, DepChecker, that assesses factuality of input text based on rules derived from structured knowledge patterns and dependency relations with respect to the input text. We show that DepChecker outperforms state-of-the-art, general purpose fact-checking techniques in this special, but important case.

We present the findings of SemEval-2022 Task 11 on Multilingual Complex Named Entity Recognition MULTICONER. Divided into 13 tracks, the task focused on methods to identify complex named entities (like names of movies, products and groups) in 11 languages in both monolingual and multi-lingual scenarios. Eleven tracks required building monolingual NER models for individual languages, one track focused on multilingual models able to work on all languages, and the last track featured code-mixed texts within any of these languages. The task is based on the MULTICONER dataset comprising of 2.3 millions instances in Bangla, Chinese, Dutch, English, Farsi, German, Hindi, Korean, Russian, Spanish, and Turkish. Results showed that methods fusing external knowledge into transformer models achieved the best results. However, identifying entities like creative works is still challenging even with external knowledge. MULTICONER was one of the most popular tasks in SemEval-2022 and it attracted 377 participants during the practice phase. 236 participants signed up for the final test phase and 55 teams submitted their systems.

Conversational Task Assistants (CTAs) are conversational agents whose goal is to help humans perform real-world tasks. CTAs can help in exploring available tasks, answering task-specific questions and guiding users through step-by-step instructions. In this work, we present Wizard of Tasks, the first corpus of such conversations in two domains: Cooking and Home Improvement. We crowd-sourced a total of 549 conversations (18,077 utterances) with an asynchronous Wizard-of-Oz setup, relying on recipes from WholeFoods Market for the cooking domain, and WikiHow articles for the home improvement domain. We present a detailed data analysis and show that the collected data can be a valuable and challenging resource for CTAs in two tasks: Intent Classification (IC) and Abstractive Question Answering (AQA). While on IC we acquired a high performing model (>85% F1), on AQA the performance is far from being satisfactory (~27% BertScore-F1), suggesting that more work is needed to solve the task of low-resource AQA.

We present AnonData, a large multilingual dataset for Named Entity Recognition that covers 3 domains (Wiki sentences, questions, and search queries) across 11 languages, as well as multilingual and code-mixing subsets. This dataset is designed to represent contemporary challenges in NER, including low-context scenarios (short and uncased text), syntactically complex entities like movie titles, and long-tail entity distributions. The 26M token dataset is compiled from public resources using techniques such as heuristic-based sentence sampling, template extraction and slotting, and machine translation. We tested the performance of two NER models on our dataset: a baseline XLM-RoBERTa model, and a state-of-the-art NER GEMNET model that leverages gazetteers. The baseline achieves moderate performance (macro-F1=54%). GEMNET, which uses gazetteers, improvement significantly (average improvement of macro-F1=+30%) and demonstrates the difficulty of our dataset. AnonData poses challenges even for large pre-trained language models, and we believe that it can help further research in building robust NER systems.

Named Entity Recognition (NER) remains difficult in real-world settings; current challenges include short texts (low context), emerging entities, and complex entities (e.g. movie names). Gazetteer features can help, but results have been mixed due to challenges with adding extra features, and a lack of realistic evaluation data. It has been shown that including gazetteer features can cause models to overuse or underuse them, leading to poor generalization. We propose GEMNET, a novel approach for gazetteer knowledge integration, including (1) a flexible Contextual Gazetteer Representation (CGR) encoder that can be fused with any word-level model; and (2) a Mixture-of- Experts gating network that overcomes the feature overuse issue by learning to conditionally combine the context and gazetteer features, instead of assigning them fixed weights. To comprehensively evaluate our approaches, we create 3 large NER datasets (24M tokens) reflecting current challenges. In an uncased setting, our methods show large gains (up to +49% F1) in recognizing difficult entities compared to existing baselines. On standard benchmarks, we achieve a new uncased SOTA on CoNLL03 and WNUT17.

The increasing popularity of voice-based personal assistants provides new opportunities for conversational recommendation. One particularly interesting area is movie recommendation, which can benefit from an open-ended interaction with the user, through a natural conversation. We explore one promising direction for conversational recommendation: mapping a conversational user, for whom there is limited or no data available, to most similar external reviewers, whose preferences are known, by representing the conversation as a user’s interest vector, and adapting collaborative filtering techniques to estimate the current user’s preferences for new movies. We call our proposed method ConvExtr (Conversational Collaborative Filtering using External Data), which 1) infers a user’s sentiment towards an entity from the conversation context, and 2) transforms the ratings of “similar” external reviewers to predict the current user’s preferences. We implement these steps by adapting contextual sentiment prediction techniques, and domain adaptation, respectively. To evaluate our method, we develop and make available a finely annotated dataset of movie recommendation conversations, which we call MovieSent. Our results demonstrate that ConvExtr can improve the accuracy of predicting users’ ratings for new movies by exploiting conversation content and external data.

Voice assistants, e.g., Alexa or Google Assistant, have dramatically improved in recent years. Supporting voice-based search, exploration, and refinement are fundamental tasks for voice assistants, and remain an open challenge. For example, when using voice to search an online shopping site, a user often needs to refine their search by some aspect or facet. This common user intent is usually available through a “filter-by” interface on online shopping websites, but is challenging to support naturally via voice, as the intent of refinements must be interpreted in the context of the original search, the initial results, and the available product catalogue facets. To our knowledge, no benchmark dataset exists for training or validating such contextual search understanding models. To bridge this gap, we introduce the first large-scale dataset of voice-based search refinements, VoiSeR, consisting of about 10,000 search refinement utterances, collected using a novel crowdsourcing task. These utterances are intended to refine a previous search, with respect to a search facet or attribute (e.g., brand, color, review rating, etc.), and are manually annotated with the specific intent. This paper reports qualitative and empirical insights into the most common and challenging types of refinements that a voice-based conversational search system must support. As we show, VoiSeR can support research in conversational query understanding, contextual user intent prediction, and other conversational search topics to facilitate the development of conversational search systems.

Recent advances in automatic speech recognition lead toward enabling a voice conversation between a human user and an intelligent virtual assistant. This provides a potential foundation for developing artificial personal shoppers for e-commerce websites, such as Alibaba, Amazon, and eBay. Personal shoppers are valuable to the on-line shops as they enhance user engagement and trust by promptly dealing with customers’ questions and concerns. Developing an artificial personal shopper requires the agent to leverage knowledge about the customer and products, while interacting with the customer in a human-like conversation. In this position paper, we motivate and describe the artificial personal shopper task, and then address a research agenda for this task by adapting and advancing existing information retrieval and natural language processing technologies.