Nai-Lung Tsao

Concept-Based Pages Recommendation by Using Cluster Algorithm

In this research, we used a proxy server to search for information related to the userpsilas browsed Web pages. From the records of the proxy server we constructed a profile of the userpsilas browsing habits. At the end of the userpsilas search subsystem, we will use content based concept to extract keywords to obtain the articlepsilas characteristicspsila description. From the recommendation system, the Web pages will be classified using the hierarchical grouping method, and through collaborative filtering, the recommendation Web pages will be chosen to provide further readings for students language learning.

CanFind-a semantic image indexing and retrieval system

We present CanFind, a semantic image indexing and retrieval system in this paper. To identify the target images of interest in the database in the conceptual level, the presented system makes use of keywords as the input of searching vehicle. The system consists of two subsystems, i.e., semantic indexing and query expansion. In the semantic indexing, the subsystem includes three main building blocks, namely, keyword extraction, keyword expansion, and keyword weighting. The information of WordNet is used to extend existing keywords associated with images. This design intends to overcome the drawbacks in conventional keyword-based image retrieval system. Next, the resulting word set is filtered by a filter to extract common words from the word set and set up the image indexing for the corresponding image. In the query expansion, corpus is used to help users find relative or precise results in the facing dilemma of too few or too many query results for a given query. The designed semantic image indexing and retrieval system is integrated with IWiLL, a web-based language learning platform to further illustrate the value of the designed system.

A ubiquitous agent for unrestricted vocabulary learning in noisy digital environments

One of the most persistently difficult aspects of vocabulary for foreign language learners is collocation. This paper describes a browser-based agent that assists learners in acquiring collocations in context during their unrestricted Web browsing. The agent overcomes the limitations imposed by learner models in traditional ITS. Its capacity to function in noisy unscripted contexts derives from a well-understood theory of lexical knowledge that attributes a words identity to its contextual features. Collocations constitute a central feature type, and we extract these features computationally from a 20-million-word portion of BNC. These we are able to detect and highlight in real time for learners in the noisy Web environments they freely browse. Our learner model, derived by semi-automatic techniques from our 3-million word corpus of learner English, maps detected collocations onto corresponding collocation errors produced by this learner population, alerting learners to the non-substitutability of words within the target collocations. A notebook offers a push function for individualized repeated exposure to examples of these collocations in context.

On the design of Web-based interactive multimedia contents for English learning

We present a Web-based authoring tool to create interactive multimedia contents for English learning. Target users of this authoring tool are English teachers and content designer, therefore, ease of use and simplicity are the fundamental issues. Furthermore, the authoring tool and multimedia database are integrated in the context of intelligent Web-based interactive language learning (IWiLL) system. Special language tools to access corpus, manipulate multimedia elements, and create collaborative learning sessions are designed in the system. These resources and tools on hand yield the potential to create rich and deep interactive multimedia contents.

Feature expansion for word sense disambiguation

One of the most serious obstacles in research on word sense disambiguation (WSD) is sparseness of training data. We describe and motivate a method of feature expansion as a means of resolving the data sparseness problem in supervised corpus-based WSD. The expanded features are extracted from an existing corpus and WordNet automatically. We use our method to supplement the feature expansion approach of [Leacock and Chodorow 1998]. In the experiments, the addition of our method more than doubled the precision improvement over baseline that was obtained by using Leacock and Chodorows approach alone.

Design and implementation of an intelligent web-based interactive language learning system

The design and implementation of an Intelligent Web-based Interactive Language Learning (IWiLL) system to support English learning on the Internet is described. We designed two kinds of learning environments: (1) interactive English writing environment and (2) mining movies for real English. These are intended to improve learners’ basic language skills such as listening, reading, and writing. In addition, the system also offers authoring tools that facilitate teachers’ content preparation. The system not only provides multimedia learning environments for users, but also builds a learner corpus, an archive of annotated English texts written by learners for whom English is a second language. Further analysis of the learner corpus creates the potential to detect the users’ persistent errors and then provide adequate help to the users.

Constructions and the problem of discovery: A case for the paradigmatic

Abstract Much of the patterned use of language occupies a poorly charted middle ground of usages that are neither frozen, one-off items listable in dictionaries nor products of maximally general rules found in grammars. Similarly, these usages fly below the radar of modular theories of language that make a strict distinction between items in a lexicon and the rules of syntax for combining them. Early constructionist approaches showed this neglected territory to be teeming with conventional form-meaning pairings, i.e., lexico-grammatical constructions. While corpora have been seen as a source for investigating these constructions, they entail a fundamental but seldom-noted limitation: constructions are constituted by both syntagmatic and paradigmatic relations, but corpora lack the paradigmatic dimension. Thus corpora can reveal multiword items but not the relations among them that constitute constructions. We elaborate on an alternative, illustrating how the design of an existing machine-readable language model affords discovery of lexico-grammatical constructions by capturing bottom up the relations they contract with other patterns. The network is noise-ridden, but we exploit the noise as the necessary background against which constructions can be set into relief and thus made discoverable.

A Time Influence Analysis for Web Page Recommender System

In this paper, we propose a time weighting model to generate a webpage recommendation list in a single website for the visitors. We use this proposed model to analyze the trend of users’ behavior, then generates a webpage recommendation list depend on it. The time weighting model provides a solution for a web page recommender system, find a better time interval to generate a webpage recommendation list, therefore enhance the recommendation precision based on a given time interval. The experimental result shows if we choose an appropriate time interval, the recommendation list generated from the specified time interval can obtain a better time weight, therefore the recommender system can make a better recommendations based on choosing an appropriate time interval for web log data mining.