Content area
Full text
1. Introduction
We introduce the novel task of understanding multi-sentence questions. Specifically, we focus our attention on multi-sentence entity-seeking questions (MSEQs), that is, the questions that expect one or more entities as answer. Such questions are commonly found in online forums, blog posts, discussion boards, etc., and come from a variety of domains including tourism, books, and consumer products.
Figure 1 shows an example of MSEQ from a tourism foruma, where the user is interested in finding a hotel that satisfies some constraints and preferences; an answer to this question is thus the name of a hotel (entity) which needs to satisfy some properties such as being a “budget” option. A preliminary analysis of such entity-seeking questions from online forums reveals that almost all of them contain multiple sentences—they often elaborate on a user’s specific situation before asking the actual question.
Fig 1.
An MSEQ annotated with our semantic labels.
[Figure omitted. See PDF]
In order to understand and answer such a user question, we convert the question into a machine representation consisting of labels identifying the informative portions in a question. We are motivated by our work’s applicability to a wide variety of domains and therefore choose not to restrict the representation to use a domain-specific vocabulary. Instead, we design an open semantic representation, inspired in part by Open QA (Fader, Zettlemoyer and Etzioni 2014), in which we explicitly annotate the answer (entity) type; other answer attributes, while identified, are not further categorized. For example, in Figure 1 “place to stay” is labeled as entity.type while “budget” is labeled as an entity.attr. We also allow attributes of the user to be represented. Domain-specific annotations such as location for tourism questions are permitted. Such labels can then be supplied to a downstream information retrieval (IR) or a QA component to directly present an answer entity.
We pose the task of understanding MSEQs as a semantic labeling (shallow parsingb) task where tokens from the question are annotated with a semantic label from our open representation. However, in contrast to related literature on semantic role labeling (SRL) (Yang and Mitchell 2017), slot-filling tasks (Bapna et al.2017), and query formulation (Vtyurina and Clarke 2016; Wang and Nyberg 2016; Nogueira and Cho 2017),...