![]() WIT Dataset is available for download and use via a Creative Commons license here. WIT represents a more diverse set of concepts and real world entities relative to what previous datasets cover. WIT is massively multilingual (first of its kind) with coverage over 100+ languages. ![]() WIT is the largest multimodal dataset by the number of image-text examples by 3x (at the time of writing). WIT is composed of a curated set of 37.5 million entity rich image-text examples with 11.5 million unique images across 108 Wikipedia languages. In this talk, I introduce the Wikipedia-based Image Text (WIT) Dataset to better facilitate multimodal, multilingual learning. ![]() ![]() Multimodal modeling techniques aim to leverage large high-quality visio-linguistic datasets for learning complementary information across image and text modalities. Wikipedia-based Image Text Dataset for Multimodal Multilingual Machine Learning By Krishna Srinivasan (Google) The milestone improvements brought about by deep representation learning and pre-training techniques have led to large performance gains across downstream NLP, IR and Vision tasks. July 2022 Time 9:30am PDT / 12:30pm EDT/ 18:30pm CEST View your local time here Theme 2022 Wikimedia Foundation Research of the Year Award Winnersǃ We would be happy to catch up in-person with you if we can align schedules. If you are in Prague in that period, feel free to ping. We are very excited that we finally can see/meet each other in person after almost 3 years of not being able to meet. The Research team will meet for an in-person offsite in Prague September 19-22. To learn more about the state of the art of Wikidata and research challenges in the era of AI/ML, we will celebrate this tenth anniversary with a panel that will bring together established researchers/practitioners in this field. The Wikimedia Research community has devoted significant effort and resources in studying the foundations, capabilities and applications of Wikidata, from the complex requirements of representing real-world knowledge in a multilingual environment to the needs to assess the quality of data and sources in Wikidata. In addition, since Wikidata is a collaborative project that can be read and edited by humans and machines alike, it is also widely used in third-party applications delivering knowledge as a service for all. The language-independent nature of Wikidata has greatly improved the maintenance and consistency of knowledge across Wikipedia language editions, fostering knowledge equity in Wikimedia. October 2022 marks the tenth anniversary of the launch of Wikidata (In ten years, this project has become the largest community-driven free knowledge graph in the world, enabling a common knowledge base for Wikimedia projects. ![]() Liaison Librarian Contribution to Local Quebecois LGBTQ+ Content in Francophone Wikipedia By Michael David Miller (McGill University)Įthical Considerations of Including Gender Information in Open Knowledge Platforms By Nerissa Lindsey (San Diego State University) Archive 2022 October 2022 Time 9:30am PDT / 12:30pm EDT / 16ː30 UTC Find your local time here Theme Panel discussion celebrating Wikidata's 10th birthdayǃīy Denny Vrandečić (WMF) with panelists Lydia Pintscher (WMDE), Elena Simperl (King's College London), Katherine Thornton (Yale), and Markus Krötzsch (Technical University of Dresden). Wikipedia and Academic Libraries By Laurie Bridges (Oregon State University) Upcoming Events November 2022 Time 9:30am PDT / 12:30pm EDT View your local time here Theme The use of open knowledge platforms by libraries We expect all presenters and attendees to abide by our Friendly Space Policy. You can join the conversation and participate in Q&A after each presentation by connecting to our IRC channel: #wikimedia-research connect. The link will be in each showcase's details below and is also announced in advance via wiki-research-l, analytics-l, and on Twitter. We live stream our research showcase every month on YouTube. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |