The document discusses topic modeling and classification of short texts. It describes using Latent Dirichlet Allocation (LDA) to extract hidden topics from a large universal text corpus consisting of Wikipedia and MEDLINE articles. These topics are then used as features for a maximum entropy classifier to categorize short texts like tweets and web snippets. Parallelized LDA is implemented using th
twitter検索 search twitter in japanese. tool:python+Django, perl, Tritonn(MySQL+Senna) Solr1.3 twitter visualization visualization of twitter public timeline.(2008/7/14 - 2008/9/7) 240,000users are sets their location in detail. tool:python, graphviz, google maps API(geocoding). HD version twitter visualization from buribunken on Vimeo
1
リリース、障害情報などのサービスのお知らせ
最新の人気エントリーの配信
処理を実行中です
j次のブックマーク
k前のブックマーク
lあとで読む
eコメント一覧を開く
oページを開く
