2006年1月24日

此感于此!

今天参观了一位同学的家。
很大的房子,很美的!
羡慕同学之余,我继续我的想法和做法。因为,每个人都有自己的梦。
祝福他们!

2006年1月23日

家[2]

在家的时间本不多,那就好好陪父母吧。
休息好了才能有更大的冲劲儿。

积蓄那份休闲,等待那份努力。
一切都那么的自然和美丽。

2006年1月22日

家是严冬下那温暖的感觉
在家很好,茫茫碌碌之中体会到的是另外的一种幸福。

2006年1月21日

[Semantic Web]ABC

[Semantic Web]ABC
--------------------------------------------
Step1: The birth May 17, 2001 on Scientific American

Title: The Semantic Web: A new form of Web content that is meaningful to computers will unleash a revolution of new possibilities
link:
English Vision
Chinese Vision

--------------------------------------------
Step 2: Primers and Guidlines

Title: The Semantic Web: A Primer
by Edd Dumbill
November 01, 2000
Link:
English Vision

Title: A Semantic Web Primer
Link:
English Vision

Totle: RDF Primer(W3C Recommendation 10 February 2004)
Link:
English Vision
Chinese Vision

Totle: Protege OWL Tutorial
Link:
English Vision

--------------------------------------------
Step3: Papers on Journals

International Journal on Semantic Web and Information Systems
Journal of Web Semantics
Web Semantics: Science, Services and Agents on the World Wide Web

--------------------------------------------
Step4: Papers on Proceedings

ISWC (International Semantic Web Conference)
WWWC(World Wide Web Conference) semantic web track
ESWS(European Semantic Web Symposium)
SWEB(International Workshop on Semantic Web Technologies in Electronic Business)
EKAW(Workshop on Knowledge Management and the Semantic Web)
SWDB(International Workshop on Semantic Web and Databases)
PPSWR(Workshop on Principles and Practice of Semantic Web Reasoning)

--------------------------------------------
Step5: Your ideas and works

……………………

2006年1月20日

My old schools

I have a soft spot in my heart for my old schools. These days, I called at my junior high school and senior high school. Some teachers had recognized me all the same. I chatted with them. There were many pieces of wonderful news about the schools. The senior high school had a website link http://www.21sjzg.com/sichuan/scedu/shool/cd/emez/index.htm. I thought now, it was very beautiful.

I loved my old schools. Because they were my education base. They kneaded and shaped my character. They teached me a lot. May they splendid achievements and futures.

2006年1月19日

Graphical Models and Inference

This afternoon, I started my winter vacation studying plan. The first material was the course Graphical Models and Inference of University of Oxford. The course was hold in October, 2005.

I had studied three chapters of it. The first three chapters were Conditional Independence, Markov properties on Undirected Graphs, and Log-Linear Models. Before my reading, I had known little about Graphical Model. Now, I had found some basic information on graphical model. To Graph Theory, there could be some probability for the node and edges. So Graphical Model had been invented. There were some famous models could be included in such theory framework, such as Markov Model, Hidden Markov Model, Maximum Entropy Markov Model, Conditional Random Field. Their relation could be realized in Graphical Model.

Althoug I had read only three chapters about it, I thouht I had been interested in it very much. I would study on it in thix vinter vocation. I would conclude in my blog after some days.

2006年1月18日

[2006-01-13]YOCSEF Conference: Content-based Searching and Search Engines

YOCSEF Conference: Content-based Searching and Search Engines
2006-01-13 14:00~18:00


This afternoon, there was an YOCSEF meeting on Content-based Searching and Search Engines. The meeting location was in Beijing University.

Our Prof. Tliu was the Executive Chairman. The three famous speakers were from Research institute of university, famous Enterprise to World-level research institute.

My supervisor Prof. Tliu gave a wonderful introduction about this meeting firstly. His speech, in my opinion, was perfect. He concluded the most important things and men of 2005.

Then Shuo Bai gave first topic on Several development on Text Mining. His presentation was wonderful. There were three main points:
1) Analysis on very huge scale rule text corpora and obtain the macroscopically characters in special scopes.
2) Model and analysis people, institute, location, events, music, software, and other abstractive objects, then obtain the related properties, relations and documents.
3) Integrate structure and un-structure data mining based on XML frames.

In his talk, I concerned some points. Document Representation has four types: Links analysis, Expand Factor (each word’s neighbors are especially. It is some statistical characters), graph expression (node is word, edge is co-occurrence frequency. Used for sub-graph finding). Document representation is the link node of natural language shallow processing, classification and clustering, and information extraction.

There were several macroscopically characters, such as single document trends analysis, attitude search engine, public opinion index, popular sequence(such as popular words, virus, and hot topic) analysis, event tracking, human tracking, and research paper&topics track.

In the audience asking time, I asked Mr. Shuo Bai a question on XML. He agreed with my opinion on XML was only a representation form and tool. The essential technology was same. To XML, document could be represented by hierarchy. With ample tools assistant, XML could be very useful to NLP and IR. Maybe it was a newly revolution.


The second topic was of Pei Chen who was the CEO of ZhongSou(Chinese Search). Pei Chen was a media event in 2005. He had several words cited often now. You could visit some of them as following:
陈沛简介
中国搜索总裁陈沛简介
中搜CEO陈沛做主题演讲
中国搜索总裁陈沛做客《专访间》
陈沛:走向中国搜索引擎4.0时代

His presentation in this forum was the future of search(搜索的未来). This was the first time I heard his presentation. In my opinion, he was sure of himself. His ZhongSou was famous now. I wished his success! He defined the third stage of search as the combination of dictionary(list earlier Yahoo) and keyword based search engine(like google just now). And his ZhongSou was ample with his idea. Yeah! When I heard his words on introducing the third stage of search, I was excited. Because I had similiar idea on our English short search engine. I was so glad to meet similiar idea. Pei Chen's analysis on the third stage of search was very good. I agreed his idea. Yes. Now we needed a stronger search engine. It should be of the feature on navigation for web and keyword searching. To the popular search engine, we could only search something. But we could not do anything for knowing things out of our mind. So this was the fault. I was looking forward more powerful one.

The third presentation was given by Dr.Huican Zhu. He was the professional engineer of Google China. He introduced some operation for business search engine. There were lots of introduction about information retrieval. I had known it a little. So I was interested in the final introduction about google. The speaker introduced the papers site of googlers. The link of google papers was http://labs.google.com/papers. I had found many good papers in this link.

Finally, Dr.Huican Zhu gave some introduction about the challenges for search engine. I recorded the later two. First was Natural Language Processing(NLP) for understanding the question and relevant facts. The other was Semantic Web which could make data easier to process and understand.

To the two challenges, NLP was our main research direction. So we could do more research on NLP for information retrieval. Semantic web had been invested a little by me. I knew it was very popular now. We could do lots of works on it.


--------------------------------------------------
In a word, this was a successful forum on searching. I gained a lot.

2006年1月17日

[Recommand]Gmail Drive: 方便外出

Gmail网络磁盘在你的电脑上利用Gmail帐号空间生成一个的网络磁盘。你可以在Gmail网络磁盘上进行任何复制、粘贴、删除、创建新文件夹,甚至拖放操作,文件存放在Gmail的服务器上,只要连着互联网,你的文件就可以随时随地拿出来分享了。当然前提是你必须有一个Gmail的帐户!

http://google.tohot.com/gmail/

2006年1月16日

Eight hours in Chengdu railway station

In the morning, at 5:30, the train arrivec at Chengdu railway station. I was excited. If everything went well, I could be home at this noon. When I went to transfer the ticket, there was a very long queue up at the box office. I could do nothing except line up in the queue.

Transport during the Spring Festival was terrible. Although it was six o'clock in the morning, there was not any ticket for my hometown in the morning. Finally, I bought the ticket at four o'clock in the afternoon. So there was ten hours for my staying at the railway station. I planed to have a good sleep there. On my way around the station, there was a Internet Bar.

I went into it and fell little warm. Because in the early morning, it was little cold.

After checking my mail box and machine learning forum, I felt very tired and sleepy. I could not help to sleeping. Before sleepiing, I opened the music of Chopin. It was after about half and one hours, I waked up. As I had had breakfast when I was waiting for the transferring ticket, I began to surf on the internet.

By google talk, I chatted with some friends with microphone. With a friend of Tianjin, we chatted about the recent research hotspot on text classification. We exchanged our ideas we learnt recently. Our conclusion was using word sense extention, we could select the better representation for document and gain the better classification performance. I told some guide from Dr. Ming Zhou to him. He suggested me to think more about the guide from Dr. Ming Zhou and choose one best direction for my Ph.D. research. Choosing is the most difficult thing in the world. I believed so.

Luckily, I met Yajie on QQ. We had missed each other four days. We talked a lot on our recent life.

In the noon, I had nice dinner in a little restaurant. I ordered Tofu pudding and Saute salted pork slice. They two were famous in Sichuan cuisine. I felt best with them.

When it was 16:10, my train started. It stopped at 18:15. When I reached in the center of my hometown, it was raining. I had not seen rain for several months. The rain was good. 19:30, I got home. My parents were excited. Me too.

2006年1月15日

Whole day on train

It was a pure whole day of my seating on train. Last night, in a word, I had not sleep well. I had only two states.

Firstly, when I was very sleepy, I armed on the corner of the table and recline my head on my arms. I could not count the time then. But maybe after half an hour, my arms were not well. I fell pins and needles in my arms. So I should change into another semi-sleeping state.

I leaned my head and back against the back of my seat and sleeped again. As you know, it was not proper pose for sleeping. I would swing my head. After standing up a short while, my head would swing into lower position. So maybe after a hour, my neck would be with pins and needles. So I should change into the first state.

As my position was in the middle and near by the passage. I had only the two sleeping states. I thought it was very unendurable. When it was nine in the morning, I had slept enough. Then I changed myself into the nice states.

In the day, I chatt little with the people in my same partition. We played card several times. Excepting chatting, sleeping, playing card, having meal, I was watching out of the windows. The scenes were chanding very fast. In the north, I could see snow. But when we went into Henan and Shanxi, there was not any snow. The sunshine was perfect. I fell very good. Yeah! If you was ln train, you would have good chance to view all kinds of scenes. It was a pleasure of journey.

2006年1月14日

By K5 for home

At 13:30, I went on the train K5 on time. It was the second train of my returning home from Harbin to Sichuan. It would a very long journey. The whole time would be 40 hours. This was the second time I seated on such long journey by train. K5 was an additional train for transport during the Spring Festival. I was little luckily that there was not so congested. We could walk to get hot water and go to WC conveniently. In term of national institute for training, today was the first day of transport during the Spring Festival. Comparing last years, this was the earliest one of my studying outbound.

How to spend the time in the 40 nhours? I had no answer firstly. I could chat and play card with others. Otherways, I could think more about my life and research. Somebody said that life is compplicated that you will do all kinds of things and appear on every occasion. On my experience, I knew seating on train was one of them. I could not change it. So why not enjoy it?

May lucky to you! If you will go for long journey in the Spring Festival, please enjoy it.

2006年1月13日

Ph.D. research topic: Guiding from Dr. Ming Zhou

This morning, I got up at 8:30. When I came to MSRA, it was about 10:00. Accompanied by Yi Chen and Jizhou Huang, I came to Dr. Ming Zhou’s office. Dr. Ming Zhou was my mentor in my first four months in MSRA. I asked some questions about Ph.D. research topics. He gave me a lot of guidance, I concluded as following:

1) To a Ph.D. research topic, you could do some applied subjects. Ph.D. subject should be very different with masters’ and undergraduates’. They should be very useful, novel, and few or no people had researched. To usefulness, it must be for practical application, such as for industrial goal or some big systems modules. Novel is same as few or no people have researched it. Yeah! If many ones have do a lot research on it, you should not do it again. Because it is time consuming and could not achieve more high level.

2) To each person, you should push over and clear yourself. If you are doing things with more old experiences, you could not be fresh and novel again. Coreference resolution is a very old research topic. Nowadays, many researchers had done lots of works on it. IBM had achieved best performance on it. If you cost three months, you will finish it. But there is no more space for your Ph.D. research.

3) Considering a Ph.D. level research topic, in the beginning, you should put yourself into a blank position. You should consider nothing especially to your old experience. The current industrial hotspots and applications would be surveyed. Then you can find out the hottest and most frontal points. Surveying such point, you could find out the one or two points for your research. Then you can start your Ph.D. research. There are two examples: Kaifu’s Statistical Speech Processing and Paraphrase in NLP.

4) Research topic should be systematical and practical. You should change your viewpoint from deeply point to whole architectural one.

I conclude Dr. Ming Zhou’s suggestion to me in three words: practical, novel, and potential. In our talking, Dr. Ming Zhou disclosed some possible research topics: Blog and RSS in Web 2.0, Multi-lingual multi-document summarization for QA, Comparison Shopping. He gave me a hint: finding research topic about Web. It was the way out.

Sincerely thanks to Dr. Ming Zhou. I will call at him again when I return to Harbin.

2006年1月12日

T158 for Beijing

Last night, after the Ph.D. Candidate Report meeting, Carl treated us a nice dinner. He was the Microsoft Scholar in 2005. We all fell exciting with him. Thanks for his nice dinner with us. Yajie took part in such dinner of our laboratory. She was welcomed. We all liked her. Then Yajie went with me to buy some food and chatted a lot with me. We said goodbye to each other before the gate of No.3 dormitory.

This morning, I went on the train T158 for Beijing. It was a long journey to us. Since the air condition was not running well. We fell sleepy in the 11 hours. Luckily, it was only 11 hours. Then Zhichang, Taozi, and me went to Train Station of West Beijing for checking transferring ticket. But there was not any proper ticket for us. We fell depressed a lot firstly. Then we decided to buy standing tickets tomorrow morning. However, we met somebody returning their tickets. We asked one by one. And luckily, we bought our tickets with seats. Taozi went on her train this evening on 1363 with sleeper. I would go on a K-type train after two days. Zhichang had bought a ticket with the help of Xu Wen. We were lucky guys.

We three had a common, expensive, happy dinner beside the train station. Then Zhichang and me sent off Taozi in the waiting room. At 10:50 she went on the train safely. Zhichang and me went to our accommodation.

In a word, it was a tired journey to us.

2006年1月11日

Ph.D. Checking report

This afternoon, the expecting Ph.D. Checking Reporting was hold. Prof. Li and Liu came here. Our ten Ph.D. Candidates reported one by one. Prof. Li and Liu gave comments after each report. Yeah! There were several students of us had achieved nice performance on many high-level papers. We all were exciting with them.

To my report, I introduced my studying state of my Ph.D. candidate. And then, there were many paper collecting information and my paper reading and research roadmaps. The final slide of my presentation was my research timeline of 2006. It’s beautiful. I thought so.

This Ph.D. Candidate Report was very important to us. We all had experienced on concluding and writing report to our works. Thanks to Prof. Li and Liu.

2006年1月10日

Ph.D. Candidate Enrollment Advice Note

I got my Ph.D. Candidate Enrollment Advice Note this afternoon. It's very exciting to me. Yeah! It was expected. At the beginning of next term, I would be a formal Ph.D. Candidate.

Study for Ph.D. degree was my biggest dream when I was young. After more than ten years diligent studying, I would stride into my dream. From enrollment to graduation, it would be a hard process. However, I love it. The result is not very important. I believe the process is perfect. I would try my best for the Ph.D. studying and do more for our IRLab.

Tomorrow afternoon, we would have the Ph.D. checking. I have prepared enough for it. There would be my research plan in 2006. Let me try more and more.

2006年1月9日

Studying Home

I will return to my hometown after several days. Now, there are many things I should finish. The most important one is defining the studying plan when I will be at home.

Yeah! I can not use computer convininently in my hometown. But it is a nice chance to studying in silence. I will bring one book and ten papers with me. They are enough for me. If more, I know I can not finish them.

Preparing some presents and printing some pictures, I have done good prepartion for my returning. Yeah, I have not seen my parents for a whole year. I am missing them very much. Tomorrow, I will prepare some other items. After the Ph.D. checking report, I will go home. Nice feeling!

2006年1月8日

Yajie on nice day

This was the last weekend which I stayed in Harbin in lunar calendar for 2005. This morning we came to look around Harbin again. This afternoon, we three had a nice dinner and went to bowling again.

From tomorrow, we all would be busy again. Nice feeling with Yajie! Thanks!

2006年1月7日

Finish the checking report

After one whole day, I finished the report for our Ph.D. checking. In the report, I concluded all my studied coures related to IR & NLP and listed my developed and researched projects in IRLab & MSRA.

Yeah! Although I had done many related works, coreference resolution which was my favorite research had not achieved any good results. After preparing more research skills and collecting the related papers, I had been up to realize my plan on coreference resolution.

In this report, I had defined my works in 2006 detaily. Five papers were in my annual plan. No more should be said. I am in the plan now.

Thanks for your concerning and help to me.

2006年1月6日

Prepare the checking report

There would be a Ph.D. candidate checking report of our lab. In our rule, I would write a report document and slides for presentation. It was a nice chance for us cleaning up all our works. I will prepare one annual plan for my research works.

2006年1月5日

IRLab New Year Conclusion and Get-Together

This afternoon, our IRLab had the annual conclusion meeting and get-together. Prof. Tliu, who was director of our lab, gave us a wonderful report. In 2005, we had done so much works and achieved nice perfoemances. But we should be progrocessing with our works also. In the beginning, Prof. Tliu gave us a good method for annual conclusion. He opened the annual plan for 2005 writing at the beginning of 2005. He checked the items one by one, analyzed the result and reasons. I thought it was a good way for conclusion.

From 15:00, we started our annual celebrating activity of our whole lab. Yiheng Chen and Shiqi Zhao gave us a wonderful celebrating activity. In the chess competition, I won once in Chinese Chess and was defeated by Xincheng Yuan. We had a lot of wonderful games. Our TM group had nice behaves. Finally, we had a good dinner.

Nice day with a new year's coming. I loved such knid conclusion meeting and activity of our lab.

2006年1月4日

Naives Bayes independent suppose

This afternoon, I worked out the experiment on Naives Bays for gender recognition. I did the experiment on w1 independent with w2. It was very important to validate my intuition.

After modifying my prior program, I had run out the final experimental result. The final whole accuracy was about 74.4%. The later experiment was on w1 and w2 which were as a whole. it was about 80.4% of whole accuracy. So the comparing experimental result told me that there was some relation between w1 and w2. I should model in detail for it tomorrow.


BTW: I'd like to study some words for describing the head of body:

auburn赤褐色的
black 黑的
blonde/golden 金色的
brown 褐色的
carroty 橘红色的
chestnut 栗色的
dark 深色的
dry 干的
dull 无光泽的
dyed 干的
fair 金色的;淡色的
flaxen 亚麻色的
greasy/oily 油性的

2006年1月3日

Professional English Writing

Although I had been writing English blogs for so long time, I believed my English was poor also. Why said so? The reason was a newly dictionary which I had brought recently.

It was an English-Chinese Dictionary of Model Compositions. In this book, there were ample categories for all kinds of description, sample sentences and sample articles. Each time, when I reviewed it, I fell my English was very poor. For example, I could not describe a person's face in my own words. And I could not write more words for snowing.

Yeah! I emphasized I should do research things in professional style. In research, writing was a very important skill. But I did not learn more on it. I wanted to learn writing in professional style. Ok. Let me read and study more on English writing. Now, I had two books on English writing: this dictionary and A Guide to Scientific Writing. In this winter vacation, I will do it.

2006年1月2日

[Recommend]Rainlendar


Rainlendar is a customizable calendar that resides on your desktop and shows the days of the current month. It's possible to add events and tasks to the calendar and the appearance can be customized with different skins. Rainlendar can also show the events and tasks from Microsoft Outlook and Mozilla Sunbird. The events can be synchronized with a server, which will allow you to use it e.g. at home and at work. There are plenty of other features too. Check the rest of this document for details.

The latest version will be found from http://www.rainlendar.net. You can also find new languages and skins from there.

More information please visit www.rainlender.net

2006年1月1日

Happy New Year's Day!

New Year, new appearance! I'd like to bless my parents, relatives, teachers, friends, classmates, IRLab members, and especially to my girl friend Yajie.

Yeah! In the past 2005, I had been learning a lot. It was of the best gain year. No more could be learnt in his year. In the first half year, I learnt more on my postgraduate courses and coreference resolution survey. I guided two undergraduates to do some basic research on centering theory and genetic algorithms. However, I changed to be intern in MSRA at May. I had not done any deep research on the two sub topics. But, in the first half year, I knew more and prepared to to more on coreference resolution research.

From May 17 to Dec. 1, I studied in MSRA which was my ideal research place before. Thanks to Prof.Tliu and Prof. Ming Zhou for giving me such a best chance. I spent four months on chatting robot design and implementation. The final two months, I changed to be guided by Dr. Cheng Niu who was very gentle and professional on information extraction. During the last period, I did some project on language independent query translation systems. I learnt more in the final two months.

From Dec. 1 to Dec. 31, I did little research works in IRLab. Although it was very short, I did lot of works. Now I knew that my working efficiency had been improved a lot comparing my situation before my interning in MSRA. I had cleared up all the research state of our coreference resolution research and collected all the papers I known. I had defined the research roadmap. The goal of our CR group was three papers at least. I will pain more.

Yeah! New year should be filled by new breath. I like such feelings.

In the past year, I'd like to list ten important things about me, as following:

1. Know and fall in love with Yajie;
2. Study in MSRA more than six months;
3. First buy presents for parents birthdays
4. Set up HIT machine learning studying group
5. Manage the skill for personal management and research group management
6. Writing blogs more in English
7. Define the roadmap for coreference resolution research
8. Master the skill on mind map
9. Know more teachers and friends: Ming Zhou, Cheng Niu, Houfeng Wang; Xiaoyuan Cui, Ke Wu, Lei Shi, Yang Zhang, Jizhou Huang, Yi Chen, and so on.
10. Be one year older than the past 2004: be more mature and know better

In 2006, I will learn more to know, lean more to do, learn more to live together, and learn more to be.