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.Documentoriented 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 documentoriented 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.

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.

In this paper, we present FigExplorer, a novel general system that supports the retrieval and exploration of research article figures. Specifically, FigExplorer can support 1) figure retrieval using keyword queries, 2) exploration of related figures of a given figure, 3) exploration of a figure topic using the citation network, and 4) search result re-ranking using an example figure. The different functions were implemented using either classical IR models or neural network-based figure embeddings. Finally, the system was designed to facilitate the collection of user data for training and test purposes and it is flexible enough such that it can be extended to include new functions and algorithms. As an open-source system, FigExplorer can help advance the research, evaluation, and development of applications in this area.

Learning to Rank is an important framework used in search engines to optimize the combination of multiple features in a single ranking function. In the existing work on learning to rank, such a ranking function is often trained on a large set of different queries to optimize the overall performance on all of them. However, the optimal parameters to combine those features are generally query-dependent, making such a strategy of "one size fits all" non-optimal. Some previous works have addressed this problem by suggesting a query-level adaptive training for learning to rank with promising results. However, previous work has not analyzed the reasons for the improvement. In this paper, we present a Best-Feature Calibration (BFC) strategy for analyzing learning to rank models and use this strategy to examine the benefit of query-level adaptive training. Our results show that the benefit of adaptive training mainly lies in the improvement of the robustness of learning to rank in cases where it does not perform as well as the best single feature.

Automated assessment of complex assignments is crucial for scaling up learning of complex skills such as critical thinking. To address this challenge, one previous work has applied supervised machine learning to automate the assessment by learning from examples of graded assignments by humans. However, in the previous work, only simple lexical features, such as words or n-grams, have been used. In this paper, we propose to use topics as features for this task, which are more interpretable than those simple lexical features and can also address polysemy and synonymy of lexical semantics. The topics can be learned automatically from the student assignment data by using a probabilistic topic model. We propose and study multiple approaches to construct topical features and to combine topical features with simple lexical features. We evaluate the proposed methods using clinical case assignments performed by veterinary medicine students. The experimental results show that topical features are generally very effective and can substantially improve performance when added on top of the lexical features. However, their effectiveness is highly sensitive to how the topics are constructed and a combination of topics constructed using multiple views of the text data works the best. Our results also show that combining the prediction results of using different types of topical features and of topical and lexical features is more effective than pooling all features together to form a larger feature space.

In this paper, we address the problem of difficult queries by using a novel strategy of collaborative query construction where the search engine would actively engage users in an iterative process to continuously revise a query. This approach can be implemented in any search engine to provide search support for users via a "Help Me Search" button, which a user can click on as needed. We focus on studying a specific collaboration strategy where the search engine and the user work together to iteratively expand a query. We propose a possible implementation for this strategy in which the system generates candidate terms by utilizing the history of interactions of the user with the system. Evaluation using a simulated user study shows the great promise of the proposed approach. We also perform a case study with three real users which further illustrates the potential effectiveness of the approach.

Enriching the content of news articles with auxiliary resources is a technique often employed by online news services to keep articles up-to-date and thereby increase users' engagement. We address the task of enriching news articles with related search queries, which are extracted from a search engine's query log. Clicking on a recommended query invokes a search session that allows the user to further explore content related to the article. We present a three-phase retrieval framework for query recommendation that incorporates various article-dependent and article-independent relevance signals. Evaluation based on an offline experiment, performed using annotations by professional editors, and a large-scale online experiment, conducted with real users, demonstrates the merits of our approach. In addition, a comprehensive analysis of our online experiment reveals interesting characteristics of the type of queries users tend to click and the nature of their interaction with the resultant search engine results page.

This work studies the effectiveness of query expansion for email search. Three state-of-the-art expansion methods are examined: 1) a global translation-based expansion model; 2) a personalized-based word embedding model; 3) the classical pseudo-relevance-feedback model. Experiments were conducted with two mail datasets extracted from a large query log of a Web mail service. Our results demonstrate the significant contribution of query expansion for measuring the similarity between the query and email messages. On the other hand, the contribution of expansion methods for a well trained learning-to-rank scoring function that exploits many relevance signals, was found to be modest.

A new class of discretization-based look-ahead algorithms (DLAAs) for the Dubins traveling salesperson problem (DTSP) is presented that compares favorably with the existing algorithms from the literature. The discretization level and the length of the look-ahead horizon are the two parameters that uniquely determine a DLAA, and depending on the application in hand, their values can be easily modified to strike a balance between the execution time and the length of the resulting admissible tour. The time complexity of a DLAA is the sum of two terms, one linear in the number of targets (cities) and one that corresponds to the specification of an initial order for the targets. For instances of the DTSP with densely distributed targets, an algorithm that relies on clustering and leads to shorter tours than the DLAA is also presented.

We present a suite of query expansion methods that are based on word embeddings. Using Word2Vec's CBOW embedding approach, applied over the entire corpus on which search is performed, we select terms that are semantically related to the query. Our methods either use the terms to expand the original query or integrate them with the effective pseudo-feedback-based relevance model. In the former case, retrieval performance is significantly better than that of using only the query, and in the latter case the performance is significantly better than that of the relevance model.