Loading Blog Search...

Wednesday, September 29, 2004

Sunday, September 26, 2004

Tips: Categorize Blogger Posts

Before read this article, I need to tell you that I have implemented a Web services to generate cateogry list for blogger. You can check out at BlogCat, which is easier to maintain than the method introduced in the following paragraphs.

This article give a tip to categorize ur posts. With tricks given, u only need to add several words in ur posts to put them into a certain category. And then, following a hyperlink, u can list all posts of a category.
Before you read this post, plz refer to another post, which bring a idea to categorize ur blogger archives. Its basic idea is to:

  1. generate a permanent link for each post on blogger, actually, this is the default action of most blogger template
  2. build some category pages that list those permanent links.
The shortcoming of this method is that u must maintain these category pages, oh, thank god, we need not to do it anymore.
  1. plz confirm whether u have generate a permanent link for each post, if no, plz follow this post to generate them.
  2. add a site search for ur blogger. I know google is the choice for most ppls, here, i recommend PICO, as i am using it. Why to use PICO? As it can index ur website when you need, and no need to wait several months for google reindexing. U donna know what's PICO site search? plz visit its homepage. :-), or try the sample in my right panel. its codes are:
    <INPUT TYPE="HIDDEN" NAME="index" VALUE="223786">
    <TD colspan="2"><A HREF="http://www.picosearch.com/cgi-bin/index.pl?wherefrom=picobox&type=all" TARGET="_blank">
    <IMG BORDER="0" SRC="http://www.picosearch.com/picosmall.gif" ALT="PicoSearch"></A>
    <INPUT TYPE="TEXT" NAME="query" VALUE="" SIZE="19"></TD>
    <tr><td><FONT SIZE=-2><A HREF="http://www.picosearch.com/cgi-bin/ts.pl?index=223786&help=help" >Help </A></FONT></td><td align=right><font size=-2><a href="http://www.picosearch.com/cgi-bin/index.pl?wherefrom=picobox&type=all" target=_blank>Site Search by PicoSearch</a></font></td></tr>
  3. add a form to ur blogger template: <form METHOD="GET" ACTION="http://www.picosearch.com/cgi-bin/ts.pl">
    <INPUT TYPE="HIDDEN" NAME="index" VALUE="223786">
    <SELECT NAME="query" width="100"><OPTION SELECTED VALUE="xxxxx Category-Research">Research   
    <OPTION VALUE="xxxxx Category-Life">Life</SELECT>
    <INPUT TYPE="SUBMIT" VALUE="List Posts" name="SEARCH">
    plz replace the index INPUT value with ur own value. Mine is 223786. Your can find it in the FORM codes given by PICO. Here, i build two categories in my blogger --- "Research" and "Life". The pith is coming. :-) plz set the value of the select input's value to some words corresponding to categories. Corresponding to research category, I set the value to "xxxxx Category-Research", where xxxxx is a word (or phrase) appearing in pemanent posts only. (I use xxxxx here is because if I use the real phrase i am using, this post will taint my category.) To life category, i set it to "xxxxx Category-Life".
  4. add the word "Category-Research" to the post, anywhere u want. Then, it is categorized to "Research" category.
  5. reindex ur pico
When writing a new post, u only need to do step 4 and 5. That's all. Now u can choose a category using the combo list, and press the button. Try the pull-down menu below. Cheers.

Wednesday, September 22, 2004

[ResearchRec]: Books read and Online Resources

Some book won't give me any appetite to have. Reversely, I cannot help buying and placing some books on my shelf. Maybe you are interested in them, maybe not. Check out what I read and some online resources I like, leave your comments if you like them too.

  • Automata, Language, Logic
    • Introduction to Automata Theory, Languages, and Computation: brought by John E. Hopcroft, Rajeev Motwani and Jeffrey D. Ullman. It's one of my textbooks in Southeast Univ. The most comprehensive one for both beginners and researhers.
    • Formal Syntax and Semantics of Programming Languages: from Ken Slonneger. I found it just now, here is its eversion. Seems interesting.
    • Introduction to Functional Programming by John Harrison, actually it's a lecture notes. this is the best one i have seen that explains how to move from lambda calcus to ML in a so simple way.

Monday, September 20, 2004

Linux Software

I found there are so many excellent tools on linux, however, always forgotten by me time and time again. So, I start this page to record my favorite tools.CategorySoft

  • Document Reader
    • CHM files: KChmPart plugin for konqueror, xchm (the only one i found print feature, still weaker than MS), gnochm.
    • MHT files: MAF plugin for firefox is good to open and save .MAF and .MHT files, unfortunately, it is only compatible with itself (not with MS). kmhtconvert is a tool convert mht to format understanded by konqueror, still in beta status.
    • PDF files: lots tools to read PDF files. pstill can concat multiple pdf files, extract pages from pdf, etc. It is free for educational purpose.
  • System Tools
    • GMailFS: We all know gmail provide 1G to each user, can u imagine how to use the 1G space? huh, mount it as a local hardisk? Here it is.
  • Document Editor
    • HTML editor: everybody may know quanta, there is a lighter WYSISWYG editor nvu
  • Entertainment
    • Video Player: MPlayer is a movie player for Linux (runs on many other Unices, and non-x86 CPUs), Windows, Mac OS X. It plays most MPEG, VOB, AVI, OGG/OGM, VIVO, ASF/WMA/WMV, QT/MOV/MP4, FLI, RM, NuppelVideo, YUV4MPEG, FILM, RoQ, PVA files, supported by many native, XAnim, and Win32 DLL codecs. You can watch VideoCD, SVCD, DVD, 3ivx, DivX 3/4/5 and even WMV movies, too (without the avifile library). PS: must known, debian does not provide official repositories for mplayer for some reasons i donno know. It's OK, here is a repository you can find mplayer for debian:
      deb ftp://ftp.nerim.net/debian-marillat/ unstable main
      To record streaming media get a URL of a stream and use command line strings like in the following examples:
          mplayer.exe -dumpstream URL
      After recording you will get the file stream.dump. You should rename it to .rm, .ra, .asf, .wmv or other format you are using. MPlayer supports HTTP, RTSP, MMS protocols to record Windows Media, RealVideo/RealAudio, QuickTime Video Recommended free MPlayer package with GUI for Windows: MPlayer + frontend by Gabor Szecsi (When using it enter your URL in a Media file editbox and "-dumpstream" (without quotes) in MPlayer Extra parameters editbox, then click "Start" button to start recording process). For other tools, plz read here.
  • Anti-Virus
      so far, i do not know who care anti-virus software on linux, although i download one just now: f-prot
  • Miscs
    • springgraph, graphviz: Springgraph is a small tool that will read in a .dot file description of a graph, which, for each node, specifies its name and which other nodes it is connected to, and then renders a graph; e.g., apt-rdepends -d springgraph >1.dot; springgraph <1.dot>1.png; will generate dependence graph of springgraph binary package. Graphviz is much more comprehensive than springgraph, and generated graph is more beautiful.
  • Graphic Editor
    • IPE: Ipe is a drawing editor for creating figures in PDF or (encapsulated) Postscript format. It supports making small figures for inclusion into LaTeX-documents as well as making multi-page PDF presentations that can be shown on-line with Acrobat Reader. So, what's the difference between IPE and xfig? I love IPE much, as it integrate latex macros seamlessly, which means u can draw tables, write formula in a convenient way (which cannot be done by xifg so far, i think).
    • QtiPlot: I used it to do my thesis graphs
    • More...更多
    • Must know, none of them has no weakness, e.g. IPE cannot import eps and pre-defined items. Another problem is although most of them can export to eps/pdf, they cannot import eps and pdf with fine effect. To combine them is a good choice, pstoedit is such a good util that can convert eps/pdf to vector format (xfig, dia, etc.) that can be used to do this.
  • Development
    • gnome source navigator: have similar features as source insight on windows, of cz, it is totally free. u can also get all reference of a function, what the function refers to, etc.
  • A quite old windows alternative software table: The table of equivalents / replacements / analogs of Windows software in Linux.

Saturday, September 18, 2004

[ResearchRec]: PageRank

Before getting pagerank ranking improvement, it's important to understand it. Of cz, it's elegant from math perspective, like HITS. Actually, it's also quite straightforward from practical perspectives. Here is an article giving some examples to explain PageRank.
Surprisely, there even exists forums focus on how to improve PageRank. Unfortunately, many people argues that google is lagging behind the changing of internet (e.g., blog is changing the internet, :-d), like yahoo!, altavista was. Here is some discussions. Concisely, I think they mean:

  • a principle property of blogs is that they are full of hyerlinks.
  • hyperlinks in blog may "taint" google's search results from perpspective of sellers and ads.
  • google must consider more about money then rank solely "popular" and adjust its PageRank.
  • of cz, the adjusted pagerank is not the pagerank we known, thus, pagerank is dead, they concluded.
The most interesting part of the "pagerank is dead" discussion is a peom from Karl Martino Pagerank is dead!? Maybe.CategoryResearch

Thursday, September 16, 2004

[ResearchRec]: Approximation algorithm

From Wikipedia, the free encyclopedia.
In computer science, approximation algorithms are an approach to attacking NP-hard optimization problems. Since it is unlikely that there can ever be efficient exact algorithms solving NP-hard problems, one settles for non-optimal solutions, but requires them to be found in polynomial time. Unlike heuristics, which just usually find reasonable good solutions reasonably fast, one wants provable solution quality and provable run time bounds. Ideally, the approximation is optimal up to a small constant factor (say 5%).

NP-hard problems vary greatly in their approximatibility; some can be approximated to arbitrary factors (such approximation algorithms are often called polynomial time approximation schemes or PTAS), some can essentially not be approximated at all.

A typical example for an approximation algorithm is the one for Vertex Cover: Find an uncovered edge and take both end points into the vertex cover. Clearly, this can only yield a set up to two times larger than the optimal one.

Frequently, one can gain approximation algorithms from examining relaxed linear programs.

Not all approximation algorithms are suitable for practical application. For example, most people would not be impressed by a scheme which provably will require them to spend less than twenty times the money that is minimally needed. Also, for some approximation algorithms, the polynomial run time can be quite bad, like O(n2000).

Another limitation of the approach is that it applies only to optimization problems and not to “pure” decision problems like Satisfiability. A compendium of NP optimization problemsCategoryResearch

Wednesday, September 15, 2004

[LifeRec]: Heavy Rain

Heavy rain in the afternoon, the world is fade behind it.

Posted by HelloCategoryLife

[LifeRec]: Diary cartoon, enjoy it...

Here is some comics updated dialy. If you want to use them, please follow hyperlinks and contact with authors. If you, the owner, do not want them to appear here, leave me a message and i will remove them.

Daily Cartoon provided by Bravenet Daily Cartoon provided by Bravenet.com
Today's Squinker Today's Mark More...

employee newsletter cartoons intranet training presentation graphics cartoon
cartoonist website humor daily comic online

Tuesday, September 14, 2004

Discussion about XML model

Some discussion
Another related interesting result is: polynomial time decidability of constraints for the ER model - this work is titled "On the Interaction between ISA and Cardinality Constraints" by D. Calvanese and M. Lenzerini, and appeared in ICDE 94. The main things about the 3 works are: 1. Henry Thompson showed that given a DTD, to check whether there is a valid instance is NP complete. It is definitely in NP, and henry showed it is NP Hard by reduction from 3SAT. 2. The work by Wenfei Fan et al, show that if we have keys and foreign keys specified as path expressions, then whether there is a valid instance that satisfies both the DTD and the constraints is undecidable (note: this is different from NPC). Note at the same time, for relational model, whether there is a valid instance for a given relational schema is trivially true. There is another result: For the relational model, whether a key is implied by a given set of keys and foreign keys is undecidable. 3. The work in ICDE 94 shows that given a ER model, whether there is a valid instance is polynomially decidable, this is true even in the presence of ISA constraints. They further show that whether a cardinality is implied by a set of ISA and cardinality constraints is decidable in polynomial time. My two cents worth: 3) is very promising. I would argue that we should design XML models so that it conforms to 3). I have tried to base our work on that - the work i pointed about constraint specification. I am not sure of the significance of trying to answer whether a key is implied by a given set of keys and foreign keys, which is undecidable even for the relational model. If you see Henry's result 1) it definitely does not conform to ER at all. More work needs to be done with respect to constraint specification. cheers and regards - murali. On XML Integrity Constraints in the Presence of DTDs Journal of the ACM (JACM), Volume 49 , Issue 3, pp 368 - 406, May 2002. Wenfei Fan and Leonid Libkin http://www.bell-labs.com/user/wenfei/papers/jacm.pdf That article should be read with care: it talks of DTDs but it means grammars in general, and it talks of integrity constraints, but these are not ID/IDREF. Thus it seems that they share the same proviso as with Henry's analysis: Henry's analysis seems to apply to DTDs with fixed IDs or fixed IDREFs, and this paper applies to where there is an outside fixed set of constraints such as keys and foreign keys. Cheers Rick Jelliffe

Interesting comparison

A guy gave a interesting comparison among three cities in american, japn, china

Thursday, September 09, 2004

Too many Gmail quote

Too many invitation right of Gmail arrived these days, donna know how to use them. If you want it, leave your email to me.

Improve traffic to your blog

In chess, promotion occurs when a pawn reaches the eighth square. At that point, your opponent is challenged because this simple pawn takes on all the powers of a Queen���the deadliest piece on the board. Of course, the pawn may be promoted to bishop, rook, or any other piece but why? This only clues the other player in to your secret plans. That being said, however, promotions to knight can often be strategically useful depending on the situation. Read more...

Wednesday, September 08, 2004

what's wrong with blogger?

its publish speed is slow, it's usually unresponse to my publish request, now, it cannot publish anymore.

About expressive capability of XML query language

XSLT, XQuery are turing-Complete. (A Simple Proof for the Turing-Completeness of XSLT and XQuery)
xpath 1 is equivalent to MSO datalog (or more general, MSO Logic).(XPath Processing in a Nutshell)
XQuery Optimization, refer to it.

Sunday, September 05, 2004

Some tips of LaTeX

To write algorithms in latex, plz use package algorithmic and algorithm, here. Plz use HEVEA, based on O'Caml to transform LaTex to HTML.

To generate tables in eps, no convenient tool there. a tip: 1. generate tables in latex; 2. latex to ps 3. ps2epsi to generate epsi files, those epsi files is ok to be changed to *.eps