. It also contains some of the best research proposals as submitted to the NLDB 2008 doctoral symposium held on June 24, 2008. The programme also includes three invited talks covering the main perspectives of the application of natural language to information systems: the way humans process, communicate and understand natural language, what are the implications and challenges towards semantic search for the new Web generation, how natural language applies to the well-established database way of querying as a means to unlock data and information for end users.We received 68 papers as regular papers for the main conference and 14 short papers for the doctoral symposium. Each paper for the main conference was assigned four reviewers based on the preferences expressed by the Program Committee members. We ensured that every paper had at least two reviewers that expressed interest in reviewing it or indicated that they could review it. We ensured that each paper got at least three reviews. As a result, only 10% of the papers were reviewed by three reviewers.The Conference Chair and the two Program Committee Co-chairs acted as Meta-Reviewers. Each of them took roughly 1/3 of the papers (obviously respecting conflicts of interest), for which s/he was responsible. This included studying the reviews, launching discussions and asking for clarifications whenever necessary, as well as studying the papers whenever a need for an informed additional opinion arose or when the reviewers' notes did not allow for a decision.After the review deadline, each meta-reviewer went through the reviews of their papers and made a preliminary assessment. To the extent possible, the meta-reviewers tried to get the inconsistencies and discrepancies resolved. We used a ranking list as resulted from the weighted average scores of all papers in a scale from 1 (lowest possible) to 6 (highest possible) as computed by taking into account the reviewer's confidence as a weighting factor. The final acceptance rate counting the number of full papers according to the NLDB tradition, at least during the last three years, was 29.4%. The following decision rules were used to make the final decisions:-Full Papers: Papers with a weighted average score of 4.0 (weak accept) or more were accepted as full papers. There were 17 such papers. One more paper was subject to conditional acceptance. This gave us room for two more papers to be accepted as full papers. We considered the papers with a weighted average score of 3.8 or above but less than 4.0 to select the two papers. Taking into account the reviewer feedback and the scores, we decided to select two papers with the least number of negative (1.0, 2.0 or 3.0) scores. -Short Papers: Papers with a weighted average score of 3.7 or above but less than 4.0 were accepted as short papers. VI Preface-Posters: Papers with a weighted average score of 3.0 (weak reject) or above but less than 3.7 were accepted as posters. -Doctoral Symposium Papers: The four-page long papers, which were invited from the submissions to the ...
This paper describes the outcomes of the First Multilingual Named Entity Challenge in Slavic Languages. The Challenge targets recognizing mentions of named entities in web documents, their normalization/lemmatization, and cross-lingual matching. The Challenge was organized in the context of the 6th Balto-Slavic Natural Language Processing Workshop, colocated with the EACL-2017 conference. Eleven teams registered for the evaluation, two of which submitted results on schedule, due to the complexity of the tasks and short time available for elaborating a solution. The reported evaluation figures reflect the relatively higher level of complexity of named entity tasks in the context of Slavic languages. Since the Challenge extends beyond the date of the publication of this paper, updates to the results of the participating systems can be found on the official web page of the Challenge.
We describe the Second Multilingual Named Entity Challenge in Slavic languages. The task is recognizing mentions of named entities in Web documents, their normalization, and cross-lingual linking. The Challenge was organized as part of the 7th Balto-Slavic Natural Language Processing Workshop, co-located with the ACL-2019 conference. Eight teams participated in the competition, which covered four languages and five entity types. Performance for the named entity recognition task reached 90% F-measure, much higher than reported in the first edition of the Challenge. Seven teams covered all four languages, and five teams participated in the cross-lingual entity linking task. Detailed evaluation information is available on the shared task web page.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.