Podcasts have shown a recent rise in popularity.Summarization of podcasts is of practical benefit to both content providers and consumers.It helps people quickly decide whether they will listen to a podcast and/or reduces the cognitive load of content providers to write summaries.Nevertheless, podcast summarization faces significant challenges including factual inconsistencies of summaries with respect to the inputs.The problem is exacerbated by speech disfluencies and recognition errors in transcripts of spoken language.In this paper, we explore a novel abstractive summarization method to alleviate these issues.Our approach learns to produce an abstractive summary while grounding summary segments in specific regions of the transcript to allow for full inspection of summary details.We conduct a series of analyses of the proposed approach on a large podcast dataset and show that the approach can achieve promising results.Grounded summaries bring clear benefits in locating the summary and transcript segments that contain inconsistent information, and hence improve summarization quality in terms of automatic and human evaluation.

In this paper, we propose a Chinese multi-turn topic-driven conversation dataset, NaturalConv, which allows the participants to chat anything they want as long as any element from the topic is mentioned and the topic shift is smooth. Our corpus contains 19.9K conversations from six domains, and 400K utterances with an average turn number of 20.1. These conversations contain in-depth discussions on related topics or widely natural transition between multiple topics. We believe either way is normal for human conversation. To facilitate the research on this corpus, we provide results of several benchmark models. Comparative results show that for this dataset, our current models are not able to provide significant improvement by introducing background knowledge/topic. Therefore, the proposed dataset should be a good benchmark for further research to evaluate the validity and naturalness of multi-turn conversation systems. Our dataset is available at https://ai.tencent.com/ailab/nlp/dialogue/#datasets.

Sentences produced by abstractive summarization systems can be ungrammatical and fail to preserve the original meanings, despite being locally fluent. In this paper we propose to remedy this problem by jointly generating a sentence and its syntactic dependency parse while performing abstraction. If generating a word can introduce an erroneous relation to the summary, the behavior must be discouraged. The proposed method thus holds promise for producing grammatical sentences and encouraging the summary to stay true-to-original. Our contributions of this work are twofold. First, we present a novel neural architecture for abstractive summarization that combines a sequential decoder with a tree-based decoder in a synchronized manner to generate a summary sentence and its syntactic parse. Secondly, we describe a novel human evaluation protocol to assess if, and to what extent, a summary remains true to its original meanings. We evaluate our method on a number of summarization datasets and demonstrate competitive results against strong baselines.

Amongst the best means to summarize is highlighting. In this paper, we aim to generate summary highlights to be overlaid on the original documents to make it easier for readers to sift through a large amount of text. The method allows summaries to be understood in context to prevent a summarizer from distorting the original meaning, of which abstractive summarizers usually fall short. In particular, we present a new method to produce self-contained highlights that are understandable on their own to avoid confusion. Our method combines determinantal point processes and deep contextualized representations to identify an optimal set of sub-sentence segments that are both important and non-redundant to form summary highlights. To demonstrate the flexibility and modeling power of our method, we conduct extensive experiments on summarization datasets. Our analysis provides evidence that highlighting is a promising avenue of research towards future summarization.

Nonrobustness is a well-known problem in many areas of computational science.Until now, robustness techniques and the construction of robust algorithms have been the province of experts in this field of research.We describe a new C/C++ library (CORE) for robust numeric and geometric computation based on the principles of Exact Geometric Computation (EGC).Through our library, for the first time, any programmer can write robust and efficient algorithms.The Core Library is based on a novel numerical core that is powerful enough to support EGC for algebraic problems.This is coupled with a simple delivery mechanism which transparently extends conventional C/C++ programs into robust codes.We are currently addressing efficiency issues in our library: (a) at the compiler and language level, (b) at the level of incorporating EGC techniques, as well as the (c) the system integration of both (a) and (b).Pilot experimental results are described.The basic library is availableathttp://cs.nyu.edu/exact/core/andtheC++-to-C compiler is under development.1 INTRODUCTION Numerical non-robustness is well-known in many areas of computational sciences and engineering [7].Nonrobustness in this paper1 refers to what is sometimes known as "catastrophic errors": errors that cause programs to enter unanticipated states and hence crash.In applications areas such as physical simulation and geometric modeling and design, the underlying geometry lWe are only interested in catastrophic errors that arise from numerical approximations.Th e computing literature often refers to preformance issues such as scalability of algorithms as "robustness issues".Such issues are also outside OUT scope.