Latent Topics in British Classic Fiction : Using LDA to classify texts into meaningful groups

エイコク　Classic　Fiction　コーパス　ノ　センザイ　テキ　トピック　LDA　ニ　ヨル　テクスト　クラスタリング

タバタ, トモジ

Tabata, Tomoji

Graduate School of Language and Culture, Osaka University

オオサカダイガク　ダイガクイン　ゲンゴブンカ　ケンキュウカ

文体

トピックモデル

共起語彙項目

意味構造

Dickens

style

topic model

lexical co-occurrence

semantic structure

The present study applies Latent Dirichlet Allocation to a corpus of British classic fiction, the Osaka Reference Corpus for Hlstrorical/Diachronic Stylistics (ORCHIDS) with a view to analyzing latent semantic structures underlying in the corpus and mapping its constituent texts in the network of words, topics, and texts. What is of special interest is that by means of this approach it is possible to shed new light on thematic structures composed by a large number of infrequent words, which would otherwise escape the net of key-word statistics due to infrequency of occurrence. Emerging results from this research are expected to lead to a new line of inquiry into key semantic patterns that characterize Dickens's semantic style as well as thematic features of 18th-and 19th-century classic fiction. The results also suggest a possibility of building a bridge between findings from machine learning text mining and stylistics, distant reading and close reading, with an empirical interplay of insights that will benefit historical/diachronic stylistics.