文章自動生成に向けた非構造データの活用の一考察
-文と文とのつながりを課題として-
1) 株式会社Speee
Abstract | ビックデータの時代が到来して数年が経過する.さらに,ここ数年の深層学習の発展は目覚しいものがあり,画像処理の分野だけでなく,自然言語処理や音声認識の分野まで及んでいる.本考察では,文章生成を実践し,そこで用いた主に3つの手法を比較考察する.1)マルコフ連鎖,2)自動要約,3)ディープラーニング(RNN/ LSTM)による文章生成.課題として,課題として,文と文とのつながりが不自然であることが検討される.実務で通用する自然な文と文とのつながりを検討する. |
---|---|
Several years have passed since the era of big data came. Furthermore, the development of deep learning in recent years has been remarkable, and it extends not only to the field of image processing but also to the field of natural language processing and speech recognition. In this study, we practice sentence generation, and mainly compare and consider the three methods used there. 1) Markov chain, 2) automatic summary, 3) sentence generation by deep learning (RNN / LSTM). As a subject, it is considered that the connection between sentence and sentence is unnatural as a subject. Consider the connection between natural sentences and sentences that are practical. | |
Keywords | 自然言語処理,非構造データ,文章自動生成 |
Natural language processing,Nonstructural data,Automatic sentence generation |