Proceedings of the First Workshop on Matching From Unstructured and Structured Data (MATCHING 2023)

Matching is a fundamental task with wide-ranging applications, encompassing search, recommendation systems, and data integration, among others. Given the proliferation of social media and e-commerce platforms, the ability to match information from both structured and unstructured sources has become increasingly crucial. At its core, matching aims to identify pairs of entries in two collections that share common properties. For instance, in the realm of HR platforms/services, matching resumes to job descriptions plays a pivotal role. Similarly, online booking platforms/services strive to match customer preferences with suitable businesses, such as hotels, restaurants, and real estate establishments. Beyond these entity matching examples, matching techniques find frequent application in various domains, such as matching customer reviews about a product to customer queries, aligning snippets of web documents with search queries, and matching user responses in Q&A platforms to new questions, etc. Consequently, matching tasks can take diverse forms based on the type of input source (structured vs. unstructured), the downstream application (eg, search, conversation, recommendation), and ethical considerations (eg, bias and transparency). The primary objective of this workshop is to facilitate collaboration between research communities in academia and industries related to several domains, including natural language processing, language generation, deep learning, conversational AI, information extraction, data integration, knowledge graphs, and human-centered computing. For this inaugural edition of the Matching Workshop, we …