1) 東京都市大学大学院
2) 東京都市大学
Abstract | 企業評価を行う上で,財務状況のみでなく,企業の戦略やその変化を捉えることが重要である.本稿では,企業の投資家向け公開情報 (IR情報) から,企業の戦略ワードを抽出することを試みる.IR情報に表れる企業の戦略ワードは造語や固有語等,企業が独自に作成したワードが多い.この様な造語や固有語は,辞書を用いた形態素解析では抽出が困難である.そこで,以下の手順で,方法を提案する.(1) N-Gramを用い戦略ワード候補の抽出.(2) 日本語に頻出するワードのコーパスをIR情報以外から作成し,ノイズの除去を行う. |
---|---|
To evaluation enterprise, it is important to consider not only enterprise financial states but also enterprise strategies or those changes. In this article, we extract the strategy words of the enterprise from the investor relations information published by the enterprise (IR information). In the enterprise strategy words appearing in IR information, there are many originally created words by companies, such as coined and proper words and so on. It is difficult to extract these words by morphological analysis with dictionary. So, we propose a method which consists of the following steps. (1) To extract candidates of strategy words by N-gram analysis. (2) To reduce noise words by using corpus of the common words that appears frequently in Japanese documents except IRs. |
|
Keywords | Nグラム,テキストマイニング,戦略ワード |
N-gram,text mining,Strategy words |