Paper Reading # 17 - Personalized Reading Support

Comments:
Comment 1
Comment 2

References:
Title: Personalized Reading Support for Second-Language Web Documents by Collective Intelligence
Authors: Yo Ehara, Nobuyuki Shimizu, Takashi Ninomiya, and Hiroshi Nakagawa
Venue: IUI 2010, Feb. 7-10 2010

Summary:
In this paper, the authors describe a new method of providing definitions for ESL readers through information gathering. Many people for whom English is the second language use programs called glossers when reading; these programs provide definitions for unfamiliar words. Most glossers automatically show the definitions for some words, and they provide this feature by choosing words that appear less frequently in the language.

With their program, they instead choose words picked by what words each individual person clicks for definitions on. They take this information and use it to calculate a person's difficulty index; that is, they determine what difficulty words they are likely to know and only gloss words that are above that difficulty. They discuss many varying algorithms that they could have used, and then show that only one method for this was suitable because it is the only one that works online. Finally, they show that the online algorithm is just as efficient as local algorithms.

Discussion:
I thought this article was interesting because I had never heard of a glosser before, and now that I have I feel like I could use one sometimes, especially when reading papers like this. The paper was very technical, which I feel is a positive, but I had a significant amount of trouble following the algorithmic analysis during the latter part of the paper. Overall however, I feel like their method would work better than previous methods and that they should make a final product out of this.

(Image courtesy of: this paper)