As a counterpart to expert-created knowledge resources such as WordNet or Cyc, non-expert users may collaboratively create large resources of unstructured or semi-structured knowledge, a leading representative of which is Wikipedia. Collectively, articles within Wikipedia form an easily-cditablc collection, reflecting an ever-growing number of topics of interest to Web users. This tutorial examines the characteristics of Wikipedia relative to other human-curated resources of knowledge; and the role of Wikipedia and resources derived from it in text analysis and in enhancing information retrieval. Applicable text analysis tasks include coreference resolution (Ratinov and Roth, 2012), word sense and entity disambiguation (Ganea and Hofmann, 2017). More prominently, they include information extraction (Zhuetal., 2019). In information retrieval, a better understanding of the structure and meaning of queries (Hu et al, 2009; Pantel and Fuxman, 2011; Tan et al., 2017) helps in matching queries against documents (Ensan and Bagheri, 2017), clustering search results (Scaiella et al., 2012), answer (Chen ct al., 2017) and entity retrieval (Ma et al., 2018) and retrieving knowledge panels for queries asking about popular entities.
展开▼