2017Q1 Reports: Conference Officer
Actions and Voices in Support of Diversity
The past year has been a memorable year for ACL to have more distinct voices and actions in support of diversity on several fronts. First, an ad hoc committee has been formed about a year ago to investigate current practices and policies for nominating and selecting various roles and awards in ACL, including ACL Fellows, life time achievements awards, and ACL chairs. A report was presented at ACL 2016 and was well received with broad interests from the community. The revised policy will be effective starting from 2017. The committee continues to work on preparing detailed recommendations and a new system to better support the execution of the new policy. Second, there have been great interests to initiate "Women in NLP" workshops in the coming NLP conferences in support of women and other minorities in NLP. Third, with leading efforts from Meg Mitchell, Emily Bender, and Hal Daume and others, ACL has made a statement in support of students and researchers from certain geographic locations whose participation may be affected in future ACL conferences.
Scaling up Conference Management in Response to Recent Growth in ACL
As there has been a considerable growth in the number of submissions made to ACL conferences in recent years, we has been increasing need to better automate core functions of the conference management system. In response to the enthusiastic request from ACL 2017 organizers and others, SoftConf has agreed to integrate Toronto Matching system to automate topic-based reviewer-to-paper assignments. The expectation is that Toronto matching system, which has been successfully used for several other conferences with considerably larger submissions, will have repeated uses for future ACL conferences. This feature complements the already existing reviewer-to-area assignments newly introduced for NAACL 2016. While the reviewer-to-area assignments work based on a manually specified set of keywords to match expertise, new Toronto matching system will match reviewers to papers based on automatically parsed information from publication records, which in turn can be automatically downloaded from people's homepages, Google Scholar, and other pages that have links to pdf papers.