ENT-DESC: Entity Description Generation by Exploring Knowledge Graph

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, pages 1187–1197, November 16–20, 2020. c⃝2020 Association for Computational Linguistics 1187 ENT-DESC: Entity Description Generation by Exploring Knowledge Graph Liying Cheng∗1,2, Dekun Wu†3, Lidong Bing2, Yan Zhang†1, Zhanming Jie†1, Wei Lu1, Luo Si2 1 Singapore University of Technology and Design 2 DAMO Academy, Alibaba Group 3 York University, Canada {liying.cheng, l.bing, luo.si}@alibaba-inc.com, jackwu@eecs.yorku.ca, {yan zhang, zhanming jie}@mymail.sutd.edu.sg, luwei@sutd.edu.sg Abstract Previous works on knowledge-to-text genera- tion take as input a few RDF triples or key- value pairs conveying the knowledge of some entities to generate a natural language descrip- tion. Existing datasets, such as WIKIBIO, WebNLG, and E2E, basically have a good alignment between an input triple/pair set and its output text. However, in practice, the input knowledge could be more than enough, since the output description may only cover the most significant knowledge. In this paper, we intro- duce a large-scale and challenging dataset to facilitate the study of such a practical scenario in KG-to-text. Our dataset involves retrieving abundant knowledge of various types of main entities from a large knowledge graph (KG), which makes the current graph-to-sequence models severely suffer from the problems of in- formation loss and parameter explosion while generating the descriptions. We address these challenges by proposing a multi-graph struc- ture that is able to represent the original graph information more comprehensively. Further- more, we also incorporate aggregation meth- ods that learn to extract the rich graph informa- tion. Extensive experiments demonstrate the effectiveness of our model architecture. 1 1 Introduction KG-to-text generation, automatically converting knowledge into comprehensive natural language, is an important task in natural language process- ing (NLP) and user interaction studies (Daml- janovic et al., 2010). Specifically, the task takes as input some structured knowledge, such as re- source description framework (RDF) triples of ∗Liying Cheng is under the Joint Ph.D. Program between Alibaba and Singapore University of Technology and Design. †Dekun Wu was a visiting student at SUTD. Yan Zhang and Zhanming Jie were interns at Alibaba. 1Our code and data are available at https://github.com/LiyingCheng95/ EntityDescriptionGeneration. Bruno�Mars retro�style,�funk,� rhythm�and�blues,� hip�hop�music,�... Peter�Gene�Hernandez�(born�October�8,�1985),�known�professionally�as Bruno� Mars,� is� an� American� singer,� songwriter,� multi-instrumentalist, record�producer,�and�dancer.�He�is�known�for�his�stage�performances, retro� showmanship� and� for� performing� in� a� wide� range� of� musical styles,�including�R&B,�funk,�pop,�soul,�reggae,�hip�hop,�and�rock.� knowledge�graph hip�hop US country�of�origin country�of citizenship genre funk Uptown�Funk performer Figure 1: An example showing our proposed task. WebNLG (Gardent et al., 2017), key-value pairs of WIKIBIO (Lebret et al., 2016) and E2E (Novikova et al., 2017), to generate natural text describing the input knowledge. In essence, the task can be formu- lated as follows: given a main entity, its one-hop at- tributes/relations (e.g., WIKIBIO and E2E), and/or multi-hop relations (e.g., WebNLG), the goal is to generate a text description of the main entity de- scribing its attributes and relations. Note that these existing datasets basically have a good alignment between an input knowledge set and its output text. Obtaining such data with good alignment could be a laborious and expensive annotation process. More importantly, in practice, the …

Content truncated to 500 words