著者

白井 寿海 十川 清己 山本 和充 児島 健次 藤岡 弘 牧田 浩和 中村 康彦

出版者

公益社団法人日本薬学会

雑誌

Biological and Pharmaceutical Bulletin (ISSN:09186158)

巻号頁・発行日

vol.16, no.2, pp.172-177, 1993

被引用文献数

7 40

---

In order to prepare fine granules of sparfloxacin (SPFX), a new quinolone anti-bacterial drug that shows masking of the bitter taste of SPFX and dissolutes at a rapid rate, various film-coated fine granules containing 20% SPFX and 0-52% low-substituted hydroxypropylcellulose(L-HPC) in the cores, were prepared by a spray method. Mixtures of ethylcellulose (EC), hydroxypropylmethylcellulose (HPMC), titanium dioxide and sucrose stearate in weight ratios of X : Y : 2 : 1 (X+Y=6) were used as film materials.The degree of masking of the bitter taste by water-insoluble film, mainly consisting of EC and HPMC, increased by increasing the content ratio of EC to HPMC and the amount of films, but was also slightly affected by the amount of L-HPC in the cores, which were coated with either EC or EC/HPMC (4/2). On the other hand, the dissolution rate increased with an increased amount of L-HPC in the cores and with a decreasing ratio of EC to HPMC in the films. Increasing the amount of L-HPC in the cores, which induced a considerable expansion of the fine granules owing to their taking up of water from the dissolution medium, resulted in burstin of the film after a short lag time. The bioavailability of the film-coated fine granules containing 20% SPFX and 52% L-HPC in the cores and 10% EC/HPMC (4/2) in the coating film, which masked the bitter taste of SPFX and showed the optimal release characteristics, was equivalent to that of conventional tablets containing 100 mg SPFX in beagle dogs.

著者

中村 康正 望月久稔

出版者

一般社団法人情報処理学会

雑誌

情報処理学会研究報告自然言語処理(NL) (ISSN:09196072)

巻号頁・発行日

vol.2005, no.94, pp.117-122, 2005-09-30

---

トライ法は,自然言語処理システムの辞書情報構築を中心に広く用いられている.このトライ法のデータ構造として,青江らが提案したダブル配列法がある.ダブル配列法は高速性とコンパクト性をあわせもっており有効なデータ構造であるが,動的検索法に比べデータの更新処理が高速であるとはいえない.そこで現在では未使用要素を単方向リストとして連結する手法が知られているが,トライ木の希点を追加および削除する際に大きなコストを必要とする.そこで本論文では,未使用要素を双方向リストとして連結することにより追加処理を高速化し,さらに削除時間を抑えるアルゴリズムを提案する.10万語の辞書データに対する実験を行った結果,追加速度は単方向リストよりも約1.5倍,削除時間は未使用要素リストを用いない従来法と同等となることが判った.A trie is used widely, such as dictionary information construction of natural language processing system. As a data structure of trie, there is the double-array structure which Aoe and others proposed. A double-array structure is an efficient data structure combining fast access with compactness. However, the updating processing is not faster than other dynamic retrieval methods. Then, although the technique of connecting empty elements as linked list is known now, big cost is needed in the node of a trie tree is inserted and deleted. In this paper, we presents a fast insertion algorithm by connecting empty elements as doubly list and reduction algorithm of deletion time. From the simulation results for 100 thousands keys, it turned out that the presented method for insertion is about 1.5 times faster than the linked list method, and deletion time is equivalent to original method which is not used 1Inked list.

著者

中村 康利

出版者

北海道大学大学院教育学研究科・教育福祉論分野

雑誌

教育福祉研究 (ISSN:09196226)

巻号頁・発行日

vol.13, pp.39-48, 2007-03-30