なかのにっき

最新版はこちら。 突っ込みは各日付の BBS エントリのほか、 メール (nakano@st.seikei.ac.jp) や フォーム からどうぞ。 なおスパム除けのため、BBS 機能には 緩い認証を入れて います。 検索エンジンから来た方は、エンジンの方のキャッシュを見るか、 下の簡易検索を試してみてください。


hns - 日記自動生成システム - Version 2.19.5

先月 2002年10月 来月
1 2 3 4 5
6 7 8 9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30 31
Namazu for hns による簡易全文検索
詳しくは 詳細指定/ヘルプを参照して下さい
検索式:

2002年10月10日(木) [n年日記]

#1 パケット落ち

家から surf までで 50% くらい drop するようだ。学内かな?

#2 [dept] 物理情報実験第三 第 1 クール 2 回目

ちょっと講義の最後でもたついた。いかん。 進行は順調。

#3 [labo] Surface Analysis 2003

メモ。

#4 [JM] mirrordir.1

あとは OPTIONS だけになったが、これがまだ 764 行もある。

#5 あちこちのサイトで... 運動

この運動の主旨は、私が理解していた限りでは、 みんなでリンクして件のページの PageRank を上げよう、というものでした *1 。「検索前にもう順番は決まってる」という、その順番自体を変えてやろう、 ということですね。 なお PageRank の再計算はたぶん結構頻繁に行なわれてると思います。 自分が登録した新規ページの Rank が、 google ツールバーに登場するまでの間隔でわかります。

さて、 PageRank 以外にもうひとつ検索に意味を持つのは、 当然のことですがページ自体がそのキーワードに一致する内容を 持っているかどうかです。 おそらくこの二つの *2 適当な重み付けによってランキングが決定されるのでしょう。

これは単に検索語がページに含まれているかどうかだけではなくて、 出現位置の html 的な文脈も計算に入っていると思います。 その間接的な証拠。 私は最初件のアンカーを <ADDRESS> 要素の内部に書いたことによって、 一時的に件のランキングのトップになってしまった のですが、これを <ADDRESS> から外してページ末尾に移動する ことによってランキングは実際に下がりました。

これを考えると、 小野澤さんの文句 は筋違いで、まず自分のページの「小野澤 英」という単語に対する一致性が悪い、 ということを認識すべき なわけです :-) だってそもそも explicit にはページのどこにも自分の名前書いてない *3 んだもん。 これを <TITLE>, <H1>, <ADDRESS> あたりに書くようにすれば、 おそらくランキングのトップに来るのはまず間違いないと思うんですけど、 まあそれはご本人が望んでいないようなので。

それ以外の今回のプロジェクトの誤算は、 件の hotwired 記事 中に書いてあった
同氏によると、今回の最も目立った調整点は、「グーグル爆撃」(googlebombing)と呼ばれる悪ふざけの影響を大幅に抑え込んだことだという。グーグル爆撃というのは、何人ものサイト運営者が自分のページ上に「go to hell」(くたばれ)というような文句書いてそれを www.microsoft.com にリンクさせておくと、グーグルで「go to hell」を検索したときに、米マイクロソフト社のホームページがトップに来る、といういたずらのテクニックだ。
にあるんじゃないかと思っています *4 。 だからリンク元のページで「小野澤 英」という単語を用いたのは、 自らのページの「小野澤 英」に対するランクを上げてしまい、 結果的に対象ページのランクを下げる効果となってしまったのでしょう。

というわけで最近の 小野澤さんの日記ページ では、末尾の「最近の一言」あたりに Webmasters が総力を結集している様子が見て取れますが、 いまのところあまり効果はあがっていないようです(笑)。
*1: 影の主旨は本人をネタにしてみんなで遊ぶこと(笑)
*2: ファクターは他にもあるかもしれませんが。
*3: 実世界でだって、 黙っていたのでは誰も自分のことを認識してくれるわけがないですね。
*4: 従って ここ での私の後段のコメントはハズレでした。

#6 [paper] 今日読んだ論文

うーん、PDF 取りに行ったときに IE5.5 が 2 回固まった。 login しっぱなしということになるので、 向こうの cookie が期限切れになるのを待たなければならないのがにんともかんとも。 IE6 だとどうなのかなあ。

Size-Dependent Detachment-Limited Decay Kinetics of Two-Dimensional TiN Islands on TiN(111):

In situ high-temperature (Ta = 1050-1250 K) scanning tunneling microscopy was used to determine the coarsening and decay kinetics of two-dimensional TiN adatom and vacancy islands on atomically smooth TiN(111) terraces. We report the first observation of an abrupt decrease in decay rates, irrespective of Ta, of adatom islands with areas less than a critical value of 1600 Angstrom^2. However, no decay rate transition was observed for vacancy islands. We attribute the size-dependent island decay behavior, which is consistent with detachment-limited kinetics, to anisotropic attachment and detachment barriers.
ふーむ、anisotoropic な系では attach-detach の energy barrier が異なる、 というのが結論だが、反応経路が違うのかな。

Organic light-emitting devices with extended operating lifetimes on plastic substrates:

有機 EL device を PET 基板上で長寿命化したというもの。 ガラスだともう数千時間の寿命、400 Cd/m^2 なんてとこまで来てるんだね。

Engineering the nanocrystalline structure of TiO2 films by aerodynamically filtered cluster deposition:

We have produced nanocrystalline titanium dioxide films with different structures (anatase or rutile) by depositing mass selected clusters from the gas phase. Nanoparticles are produced by a pulsed microplasma cluster source and are selected by aerodynamic separation effects. We have characterized nanocrystalline films by Raman spectromicroscopy and transmission electron microscopy, showing that the films assembled with very small clusters have a predominant rutile phase, whereas larger clusters form films with anatase structure. Our observations suggest that phonon confinement effects are responsible for a significant shift and broadening observed for the Raman peaks.

Stressed metal probes for atomic force microscopy:

こ、これは...(^^;

Sputtering from ion-beam-roughened Cu surfaces:

A comprehensive theoretical and experimental study of sputtering from copper surfaces roughened by low-energy Ar+ ion bombardment is reported. The total sputtering yields of thermally deposited Cu samples bombarded by 400-eV and 800-eV ions at 0°-70° angles of incidence have been measured and compared with a numerical model we have developed. To compute sputtering yields from rough surfaces, an original approach has been introduced, which accounts for sputtering anisotropy and shadowing of material emitted at grazing angles. The approach is flexible with respect to surface morphology and can be applied with any submicron structures. To specify the morphology that develops on the Cu surface under low-energy ion bombardment, the surface of bombarded Cu samples has been investigated by scanning electron microscopy. The morphology has been found highly unstable, appearing with random roughening, inclined conelike structures, ripples, or almost flat surfaces, depending on the bombardment conditions. For the samples considered it is found that the angular dependency of the total sputtering yield is strongly affected by surface morphology, which varies with the angle of ion incidence and bombardment energy. Approximations for accounting for the surface roughness required to describe sputtering at particular energy and angular regimes are discussed.
おお、すばらしい。って Dew 先生か。 モデルはかなり単純だけど、実測結果が入ってるのがありがたい。

#7 [CD] 『檸檬』遊佐未森

「ゴンドラの唄」とか「蘇州夜曲」とか。声質にあってて良い〜

#8 [dept] 発注記録

河田先生からの電話を受けて、 藤本理化に一級 KCl 500g を 2 本注文。 到着したらメールで連絡のこと。
コメント [全部読む/投稿する]

以上、1 日分です。
タイトル一覧
カテゴリ分類
book
dept
issp
labo
paper
snap
stock
vsj
Powered by hns-2.19.5, HyperNikkiSystem Project

中野武雄 (NAKANO, Takeo) <nakano@st.seikei.ac.jp> Since 1999-10-07
RSS feed, 更新時刻, LIRS エントリ, アクセス制御 (解説)

中野のホームページへ