xuy1202
Dec 17, 2009
Graduate / Professional Statement for CS, Information retrieval [3]
Thank you so much indeed!
nothing escapes you!
Thank you so much indeed!
nothing escapes you!
xuy1202
Dec 16, 2009
Graduate / Professional Statement for CS, Information retrieval [3]
My research interest regarding information retrieval derives from "The National Science and Technology Service Network" project, in which I was in charge of the development and maintenance of information classification and retrieval. The project taught me that special information is needed by individual users; however, the transfer process was impeded by the absence of effective search engine optimization and the vast amount of information noise. This is common with big websites. Considering the amount of information being continually published on the Internet, it is nearly impossible to collect all the information required by any user, especially if the user knows little about computer technology and advanced search methods. So, effective retrieval of information aroused my interest.
Meanwhile, another project, "The Evaluation of Contribution Power of Chinese Agriculture Website Information," which analyzes the Chinese agricultural website, demonstrated the challenge of finding information even if web sites have a lot of it. Not only is it difficult to become familiar with all those websites, but it is also time-consuming to seek information on web sites one by one.
Therefore, I chose the agricultural information-orientated intelligent meta-search engine as my graduate project, with the goal of developing an intelligent web system to collect useful information from several hundreds of agricultural web sites to produce a more effective way to gather information. This technology is especially useful for filling the information gap in China, but can be used anywhere.
Through research of the meta-search engine, I realized that the accuracy of any evaluation methods and relevance algorithm pales in comparison with the user's judgment. With that in mind, I designed a user search behavior-based method to deal with the results cluster. Lab experiments proved its effectiveness and efficiency. Moreover, compared with the demand of a high recall ratio, the search engine, in my mind, should be deemed effective as long as it can offer the most-needed results. The so-called comprehensive results, due to replication, are meaningless repetition, wasting the system's resources and the user's time. Considering that different users utilize the same key words to search for different results, identifying the limited results set that can satisfy the largest proportion of users' needs is the key point. Currently, I am trying to design a user feedback-based sub-engine scheduling method in order to determine the most useful results set through continuous interaction of user searches and results chosen.
In the future, to assist my research in information retrieval, I want to broaden my academic knowledge about machine learning (ML), human computer interaction (HCI), information visualization (IV) or even cognitive psychology (CP). I want to devote my time to studying the interactive process between user and system: how to use HCI to obtain the information about user's true needs, how to use ML to improve the IR model and method and how to use IV to organize information results for a better results display.
As my research progresses, I feel the pain of not finding an immediate resolution to the problem and the happiness of inspiration burst. However, the pain leads to personal growth and resolve, while the happiness, which is like the lively cabin, smoke dancing from its kitchen chimney, I want to build if I get stuck on a deserted island, led me to make a decision to dedicate my life to research. Acquisition of "things" doesn't make me happy; creation does.
Although my masters program is near completion and I have learned a lot, my research about information retrieval is just beginning. I want to further my study through a PhD degree. I am familiar with ABC University and the PhD program from the Internet, and I believe it is the right place for me to accomplish my goals. Moreover, I believe that my analytical ability, laboratory experience and theoretical knowledge gained through research and advanced courses can help me excel in a rigorous doctoral program and allow me to make a contribution to the program as well.
-----------------------------------------------
appreciate your suggestion : )
My research interest regarding information retrieval derives from "The National Science and Technology Service Network" project, in which I was in charge of the development and maintenance of information classification and retrieval. The project taught me that special information is needed by individual users; however, the transfer process was impeded by the absence of effective search engine optimization and the vast amount of information noise. This is common with big websites. Considering the amount of information being continually published on the Internet, it is nearly impossible to collect all the information required by any user, especially if the user knows little about computer technology and advanced search methods. So, effective retrieval of information aroused my interest.
Meanwhile, another project, "The Evaluation of Contribution Power of Chinese Agriculture Website Information," which analyzes the Chinese agricultural website, demonstrated the challenge of finding information even if web sites have a lot of it. Not only is it difficult to become familiar with all those websites, but it is also time-consuming to seek information on web sites one by one.
Therefore, I chose the agricultural information-orientated intelligent meta-search engine as my graduate project, with the goal of developing an intelligent web system to collect useful information from several hundreds of agricultural web sites to produce a more effective way to gather information. This technology is especially useful for filling the information gap in China, but can be used anywhere.
Through research of the meta-search engine, I realized that the accuracy of any evaluation methods and relevance algorithm pales in comparison with the user's judgment. With that in mind, I designed a user search behavior-based method to deal with the results cluster. Lab experiments proved its effectiveness and efficiency. Moreover, compared with the demand of a high recall ratio, the search engine, in my mind, should be deemed effective as long as it can offer the most-needed results. The so-called comprehensive results, due to replication, are meaningless repetition, wasting the system's resources and the user's time. Considering that different users utilize the same key words to search for different results, identifying the limited results set that can satisfy the largest proportion of users' needs is the key point. Currently, I am trying to design a user feedback-based sub-engine scheduling method in order to determine the most useful results set through continuous interaction of user searches and results chosen.
In the future, to assist my research in information retrieval, I want to broaden my academic knowledge about machine learning (ML), human computer interaction (HCI), information visualization (IV) or even cognitive psychology (CP). I want to devote my time to studying the interactive process between user and system: how to use HCI to obtain the information about user's true needs, how to use ML to improve the IR model and method and how to use IV to organize information results for a better results display.
As my research progresses, I feel the pain of not finding an immediate resolution to the problem and the happiness of inspiration burst. However, the pain leads to personal growth and resolve, while the happiness, which is like the lively cabin, smoke dancing from its kitchen chimney, I want to build if I get stuck on a deserted island, led me to make a decision to dedicate my life to research. Acquisition of "things" doesn't make me happy; creation does.
Although my masters program is near completion and I have learned a lot, my research about information retrieval is just beginning. I want to further my study through a PhD degree. I am familiar with ABC University and the PhD program from the Internet, and I believe it is the right place for me to accomplish my goals. Moreover, I believe that my analytical ability, laboratory experience and theoretical knowledge gained through research and advanced courses can help me excel in a rigorous doctoral program and allow me to make a contribution to the program as well.
-----------------------------------------------
appreciate your suggestion : )