Classifying Wikipedia entities into fine-grained classes

Classifying Articles in Chinese Wikipedia with Fine-Grained Named Entity Types

“…For different languages, it is difficult to perfectly implement the previous works because language-dependent features and heuristic rules are usually adopted to achieve better classification performance, such as in Japanese [19] and Arabic [20]. In order to evaluate the resulting training set in classification, we re-implemented the state-of-the-art method presented by Tkatchenko et al [17], and their baseline method that is similar to Tardif's classifier [18]. Their baseline method used the text of the first paragraph as a basic feature space, and a range of additional ones, namely, Title, Infobox, Sidebar, and Taxobox tokens, stemmed and tokenized categories, and template names.…”

“…Saleh et al [16] extracted features from abstracts, infobox, category, and persondata structure, and improved the recall of different NE types, by using beta-gamma threshold adjustment. Tkatchenko et al [17] adopted similar features to Tardif et al [18]. They added a 'List of' feature to the bag-of-words (BOW) representation, and added a boolean feature, which is the result of a binary rule, to increase separability between the articles of NEs and non-entities.…”

“…These hierarchical layouts, especially list items, are usually used to extract hyponymy-relation candidates and to then realize large-scale hyponymy-relation acquisition [22,23]. Tkatchenko et al [17] added the 'List of' feature, which is constructed by tokenizing and stemming the titles of all 'List of …' articles containing the current article, to the BOW representation. In this paper, the co-occurrence articles are utilized to extract representative words and increase their weights because only a small number of articles tagged with co-occurrence relation can be found in Chinese Wikipedia.…”

Identificación de relaciones entre los nodos de una red social

“…También Bollegala [38] realiza estudios sobre la identifi cación de entidades nombradas en microtextos o textos breves que se publican en redes sociales como Twitter y Facebook. Otros estudios como el de Tkatchenko [39] propone un enfoque semi-supervisado para la construcción de conjuntos de entrenamiento para la clasifi cación de entidades nombradas. Para su desarrollo se usó una taxonomía de entidades nombradas llamada BBN [40] , un umbral de al menos 40 artículos de Wikipedia, y un subconjunto de las 400 palabras en minúscula más frecuentes, del corpus Reuters.…”

“…En 2007, Hirano [51] propone adicionar un mecanismo de aprendizaje supervisado al proceso, el cual mejora en un 4.4% la precisión. En 2011, Tkatchenko [39] propone utilizar un clasifi cador con aprendizaje semi-supervisado basado en SVM para establecer relaciones entre entidades nombradas dentro de Wikipedia. El trabajo ofrece niveles de precisión cercanos a 1 (100%) al aplicar el clasifi cador sobre 18 clases, lo cual es un resultado destacable.…”