Generating Question-Answer Hierarchies

Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 2321–2334 Florence, Italy, July 28 - August 2, 2019. c⃝2019 Association for Computational Linguistics 2321 Generating Question-Answer Hierarchies Kalpesh Krishna & Mohit Iyyer College of Information and Computer Sciences University of Massachusetts Amherst {kalpesh,miyyer}@cs.umass.edu http://squash.cs.umass.edu/ Abstract The process of knowledge acquisition can be viewed as a question-answer game between a student and a teacher in which the student typi- cally starts by asking broad, open-ended ques- tions before drilling down into specifics (Hin- tikka, 1981; Hakkarainen and Sintonen, 2002). This pedagogical perspective motivates a new way of representing documents. In this paper, we present SQUASH (Specificity-controlled Question-Answer Hierarchies), a novel and challenging text generation task that con- verts an input document into a hierarchy of question-answer pairs. Users can click on high-level questions (e.g., “Why did Frodo leave the Fellowship?”) to reveal related but more specific questions (e.g., “Who did Frodo leave with?”). Using a question taxonomy loosely based on Lehnert (1978), we classify questions in existing reading comprehension datasets as either GENERAL or SPECIFIC. We then use these labels as input to a pipelined system centered around a conditional neu- ral language model. We extensively evaluate the quality of the generated QA hierarchies through crowdsourced experiments and report strong empirical results. 1 Introduction Q: What is this paper about? A: We present a novel text generation task which converts an input document into a model- generated hierarchy of question-answer (QA) pairs arranged in a top-down tree structure (Fig- ure 1). Questions at higher levels of the tree are broad and open-ended while questions at lower levels ask about more specific factoids. An entire document has multiple root nodes (“key ideas”) that unfold into a forest of question trees. While readers are initially shown only the root nodes of the question trees, they can “browse” the document by clicking on root nodes of interest Massive Attack (band) On 21 January 2016, the iPhone application "Fantom" was released. The application was developed by a team including Massive Attack's Robert Del Naja and let users hear parts of four new songs by remixing them in real time, using the phone's location, movement, clock, heartbeat, and camera. On 28 January 2016, Massive Attack released a new EP, Ritual Spirit, which includes the four songs released on Fantom. Q. What was the iPhone application Fantom? A. The app... let users hear parts of ... real time, Q. Who created it? A. ... team including ... Robert Del Naja On 26 July 2016, Massive Attack previewed three new songs: "Come Near Me", "The Spoils", and "Dear Friend" on Fantom, an iPhone application on which they previously previewed the four songs from the Ritual Spirit EP ... The video for "The Spoils", featuring Cate Blanchett, and directed by Australian director John Hillcoat, ... Q. What is Ritual Spirit? A. On ... Attack released a new EP Ritual Spirit Q. What did they do in 2016? A. On ... 2016 ... three new songs: …

Content truncated to 500 words