Entity disambiguation (ED) is the last step of entity linking (EL), when candidate entities are reranked according to the context they appear in. All datasets for training and evaluating models for EL consist of convenience samples, such as news arti
cles and tweets, that propagate the prior probability bias of the entity distribution towards more frequently occurring entities. It was previously shown that the performance of the EL systems on such datasets is overestimated since it is possible to obtain higher accuracy scores by merely learning the prior. To provide a more adequate evaluation benchmark, we introduce the ShadowLink dataset, which includes 16K short text snippets annotated with entity mentions. We evaluate and report the performance of popular EL systems on the ShadowLink benchmark. The results show a considerable difference in accuracy between more and less common entities for all of the EL systems under evaluation, demonstrating the effects of prior probability bias and entity overshadowing.
Citing legal opinions is a key part of legal argumentation, an expert task that requires retrieval, extraction and summarization of information from court decisions. The identification of legally salient parts in an opinion for the purpose of citatio
n may be seen as a domain-specific formulation of a highlight extraction or passage retrieval task. As similar tasks in other domains such as web search show significant attention and improvement, progress in the legal domain is hindered by the lack of resources for training and evaluation. This paper presents a new dataset that consists of the citation graph of court opinions, which cite previously published court opinions in support of their arguments. In particular, we focus on the verbatim quotes, i.e., where the text of the original opinion is directly reused. With this approach, we explain the relative importance of different text spans of a court opinion by showcasing their usage in citations, and measuring their contribution to the relations between opinions in the citation graph. We release VerbCL, a large-scale dataset derived from CourtListener and introduce the task of highlight extraction as a single-document summarization task based on the citation graph establishing the first baseline results for this task on the VerbCL dataset.
We analyse the performance of passage retrieval models in the presence of complex (multi-hop) questions to provide a better understanding of how retrieval systems behave when multiple hops of reasoning are needed. In simple open-domain question answe
ring (QA), dense passage retrieval has become one of the standard approaches for retrieving the relevant passages to infer an answer. Recently, dense passage retrieval also achieved state-of-the-art results in multi-hop QA, where aggregating information from multiple documents and reasoning over them is required. However, so far, the dense retrieval models are not evaluated properly concerning the multi-hop nature of the problem: models are typically evaluated by the end result of the retrieval pipeline, which leaves unclear where their success lies. In this work, we provide an in-depth evaluation of such models not only unveiling the reasons behind their success but also their limitations. Moreover, we introduce a hybrid (lexical and dense) retrieval approach that is highly competitive with the state-of-the-art dense retrieval model, while requiring substantially less computational resources. Furthermore, we also perform qualitative analysis to better understand the challenges behind passage retrieval for multi-hop QA.
Conversational search is a relatively young area of research that aims at automating an information-seeking dialogue. In this paper we help to position it with respect to other research areas within conversational Artificial Intelligence (AI) by anal
ysing the structural properties of an information-seeking dialogue. To this end, we perform a large-scale dialogue analysis of more than 150K transcripts from 16 publicly available dialogue datasets. These datasets were collected to inform different dialogue-based tasks including conversational search. We extract different patterns of mixed initiative from these dialogue transcripts and use them to compare dialogues of different types. Moreover, we contrast the patterns found in information-seeking dialogues that are being used for research purposes with the patterns found in virtual reference interviews that were conducted by professional librarians. The insights we provide (1) establish close relations between conversational search and other conversational AI tasks; and (2) uncover limitations of existing conversational datasets to inform future data collection tasks.
This paper describes the participation of UvA.ILPS group at the TREC CAsT 2020 track. Our passage retrieval pipeline consists of (i) an initial retrieval module that uses BM25, and (ii) a re-ranking module that combines the score of a BERT ranking mo
del with the score of a machine comprehension model adjusted for passage retrieval. An important challenge in conversational passage retrieval is that queries are often under-specified. Thus, we perform query resolution, that is, add missing context from the conversation history to the current turn query using QuReTeC, a term classification query resolution model. We show that our best automatic and manual runs outperform the corresponding median runs by a large margin.
Conversational passage retrieval relies on question rewriting to modify the original question so that it no longer depends on the conversation history. Several methods for question rewriting have recently been proposed, but they were compared under d
ifferent retrieval pipelines. We bridge this gap by thoroughly evaluating those question rewriting methods on the TREC CAsT 2019 and 2020 datasets under the same retrieval pipeline. We analyze the effect of different types of question rewriting methods on retrieval performance and show that by combining question rewriting methods of different types we can achieve state-of-the-art performance on both datasets.
How can we better understand the mechanisms behind multi-turn information seeking dialogues? How can we use these insights to design a dialogue system that does not require explicit query formulation upfront as in question answering? To answer these
questions, we collected observations of human participants performing a similar task to obtain inspiration for the system design. Then, we studied the structure of conversations that occurred in these settings and used the resulting insights to develop a grounded theory, design and evaluate a first system prototype. Evaluation results show that our approach is effective and can complement query-based information retrieval approaches. We contribute new insights about information-seeking behavior by analyzing and providing automated support for a type of information-seeking strategy that is effective when the clarity of the information need and familiarity with the collection content are low.
The dependency between an adequate question formulation and correct answer selection is a very intriguing but still underexplored area. In this paper, we show that question rewriting (QR) of the conversational context allows to shed more light on thi
s phenomenon and also use it to evaluate robustness of different answer selection approaches. We introduce a simple framework that enables an automated analysis of the conversational question answering (QA) performance using question rewrites, and present the results of this analysis on the TREC CAsT and QuAC (CANARD) datasets. Our experiments uncover sensitivity to question formulation of the popular state-of-the-art models for reading comprehension and passage ranking. Our results demonstrate that the reading comprehension model is insensitive to question formulation, while the passage ranking changes dramatically with a little variation in the input question. The benefit of QR is that it allows us to pinpoint and group such cases automatically. We show how to use this methodology to verify whether QA models are really learning the task or just finding shortcuts in the dataset, and better understand the frequent types of error they make.
We introduce a new dataset for Question Rewriting in Conversational Context (QReCC), which contains 14K conversations with 80K question-answer pairs. The task in QReCC is to find answers to conversational questions within a collection of 10M web page
s (split into 54M passages). Answers to questions in the same conversation may be distributed across several web pages. QReCC provides annotations that allow us to train and evaluate individual subtasks of question rewriting, passage retrieval and reading comprehension required for the end-to-end conversational question answering (QA) task. We report the effectiveness of a strong baseline approach that combines the state-of-the-art model for question rewriting, and competitive models for open-domain QA. Our results set the first baseline for the QReCC dataset with F1 of 19.10, compared to the human upper bound of 75.45, indicating the difficulty of the setup and a large room for improvement.
The ability to engage in mixed-initiative interaction is one of the core requirements for a conversational search system. How to achieve this is poorly understood. We propose a set of unsupervised metrics, termed ConversationShape, that highlights th
e role each of the conversation participants plays by comparing the distribution of vocabulary and utterance types. Using ConversationShape as a lens, we take a closer look at several conversational search datasets and compare them with other dialogue datasets to better understand the types of dialogue interaction they represent, either driven by the information seeker or the assistant. We discover that deviations from the ConversationShape of a human-human dialogue of the same type is predictive of the quality of a human-machine dialogue.
