Towards Machine Reading for Interventions from Humanitarian-Assistance Program Literature

Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing, pages 6444–6448, Hong Kong, China, November 3–7, 2019. c⃝2019 Association for Computational Linguistics 6444 Towards Machine Reading for Interventions from Humanitarian-Assistance Program Literature Bonan Min, Yee Seng Chan, Haoling Qiu, and Joshua Fasching Raytheon BBN Technologies, Cambridge, Massachusetts {bonan.min,yeeseng.chan,haoling.qiu,joshua.fasching}@raytheon.com Abstract Solving long-lasting problems such as food insecurity requires a comprehensive under- standing of interventions applied by govern- ments and international humanitarian assis- tance organizations, and their results and con- sequences. Towards achieving this grand goal, a crucial first step is to extract past interven- tions and when and where they have been ap- plied, from hundreds of thousands of reports automatically. In this paper, we developed a corpus annotated with interventions to foster research, and developed an information extrac- tion system for extracting interventions and their location and time from text. We demon- strate early, very encouraging results on ex- tracting interventions. 1 Introduction The world is a complex socio-political system: there are long lasting problems such as food in- security, global warming and diseases affecting much of the world’s population, as well as burst- ing, extreme events such as war, natural disas- ters and financial crisis. Recently, there has been growing interests in applying event extraction to provide better situation awareness (“what hap- pened”), but rarely in terms of offering insight into providing guidance on what humanitarian assis- tance interventions have been applied in the past and how they influence the situation. Furthermore, interventions may have intended outcomes and unintended consequences. An ex- ample of an unintended consequence is that free food distribution depresses prices for local pro- duce, and creates a disincentive for farmers. It is extremely useful to automatically extract interven- tions, their outcome and consequences from hun- dreds of thousands of articles, to provide to deci- sion makers in governments or humanitarian aid organizations a comprehensive understanding of what intervention options are available when cri- sis happened, and what outcomes to expect when applying each of them. This paper is a first step in this direction, starting with developing Information Extraction (IE) techniques to automatically extract interven- tions from text including academic studies, pro- gram/project guidance and evaluation documents from non-profit or international organizations. We view interventions as a (series of) event(s) 1 with time and space dimensions. For example: S1: WFP is scaling up its food assistance activ- ities in Baghdad, Anbar, Dohuk and Ninewa gov- ernorates in 2018. The intervention “Food Assistance” is happen- ing in the year 2018 in locations “Baghdad, Anbar, Dohuk and Ninewa governorates”. Being able to extract interventions and their location and time is a first step towards enabling comprehensive under- standing of their effects and consequences. In this paper, we develop IE techniques towards reading for interventions. The basic methodology is to treat intervention extraction as an event ex- traction problem, and read intentional and factual statements about interventions in existing project documentation and evaluations literature. Our …

Content truncated to 500 words