With advances in network technologies, geographic barriers are hardly a problem now for global communication. Languages, whether written or spoken, are the major tools for cyber communication, and English, with its wide popularity, has been recognized as a global language. For non-native English- speaking people, extensive reading is a common way to improve a person's command of English. English is even taken as a major course in primary schools in many countries where English as a Foreign Language (EFL) is taught, especially East Asia. One of the keys to success in English learning depends on a person's vocabulary volume. Improving the English vocabulary of a learner has thus become a popular research issue in countries where EFL is widely taught.
English teaching in Taiwan usually emphasizes the analysis and memorization of stems, prefixes, and suffixes of unknown words and uses vocabulary in accordance with a learner's command of English to explain the meaning of these words. This practice enables a learner to remember the meaning of a word, but it may not allow the learner comprehend the word and be able to use the word in different circumstances. As such, the learner may forget it quickly. To build up one's vocabulary volume quickly and sustainably, the best method is not to remember the words by rote but rather to read extensively and often. These experts all suggest that a long-term habit of extensively reading articles that are appropriate for a learner's English ability can greatly improve the vocabulary and command of a learner of English (Song, 2000; Xuan, 2002; Chen & Hsu; 2008a). However, this strategy may be difficult to implement for a learner with no extensive vocabulary because the learner may have problems either in choosing appropriate levels of articles in accordance with her/his needs and interests or in figuring out the meaning of unknown words using the semantics of familiar words in an article that is obtained. Dictionaries are always helpful; however, the need to continually look up unknown words, which once learned may be forgotten in a few days as per learning curve theory, might also easily discourage a learner. Several studies have also developed language tutoring systems in order to assist learners in learning language (Heift & Nicholson, 2001; Hsu, 2008; Ferreira & Atkinson, 2009) and Essalmi et al. (2010) have also proposed different personalization strategies for personalized e-learning systems by using 18 personalization parameters. With regard to an e-learning system, it is also hard to select appropriate articles for such learners if details on or precise profiles of learners are never established.
This paper proposes an approach that can help a learner build up his/her English vocabulary volume by intensive article reading, during which meanings of unknown words are understood in the context of articles, paired with immediate tests, through which learners can further understand the vocabulary in the articles as well as can enhance the suitability of article recommendations. Beginners in English learning usually know only a limited number of vocabulary and phrases; it would be quite hard for them to figure out meanings of unknown words in an article beyond their ability. The proposed approach uses modern item response theory (IRT) to recursively calculate the current vocabulary ability of a learner and harnesses fuzzy logic and memory cycle theories to choose suitable articles from a pool of graded articles for the learner. After an article is read, the vocabulary in it, which a learner is supposed to have comprehended, is used to generate a test in order to check the degree to which the learner understands the article by focusing on vocabulary and phrases. A prototype system was built, and several experiments were conducted to evaluate the efficiency of the proposed approach. Analysis of learner achievements and questionnaires confirm that the proposed approach not only enhances the English ability learners but also enhances their learning interest as an effect of adaptive learning. …