In this paper we show how to learn rules to improve the performance of a machine translation system. Given a system consisting of two translation functions (one from language A to language B and one from B to A), training text is translated from A to B and back again to A. Using these two translations, differences in knowledge between the two translation functions are identified, and rules are ...

Introduction. This paper provides a contribution to understandings of information literacy regarding context and transferability of information practices. Specifically, the paper analyses the subset of information practices in situations of student protests and addresses issues of transfer of information literacy practice from a highly formal landscape to a highly non-formal and unstructured la...

This paper revisits some foundational questions concerning the abstract representation of a discourse context. The context of a conversation is represented by a body of information that is presumed to be shared by the participants in the conversation – the information that the speaker presupposes a point at which a speech act is interpreted. This notion is designed to represent both the informa...

In this paper, we provide a comprehensive survey of the space of possible analyses of the phenomenon of quantifier domain restriction, together with a set of considerations which militate against all but our own proposal. Among the many accounts we consider and reject are the ‘explicit’ approach to quantifier domain restriction discussed, for example, by Stephen Neale, and the pragmatic approac...

This paper presents an approach for representing, and providing computer support for, the configuration of interactive systems, particularly ubiquitous systems, that offers a flexible method for combining a wide range of configuration techniques. There are many existing techniques offering dynamic adaptation, ranging from fully automatic through context-sensitive to userdriven. We propose a mod...

あらまし 分散情報表現を用いた神経回路網において,同じ入力から文脈に応じて異なるパターンを連想する 際には,主に入力パターンに文脈を表すパターンを連結する方法が用いられてきた.しかし,この方法には本質 的な問題点があり,入力及び文脈パターンの数に関して強い制約が生じていた.本論文では,これとは異なる文 脈修飾の方式を非単調神経回路網に適用することによって,従来の問題点を解決する文脈依存的連想モデルを構 築する.このモデルでは,入力及び文脈パターン数に関係なく,学習できる連合の数が素子数にほぼ比例して増 加する.また,文脈を切り換えることによって,アトラクタ間の状態遷移を自在に制御することができる.本モ デルは,素子数や学習時間の爆発を招くことなく任意の有限オートマトンの動作を模擬可能であると同時に,完 全な分散表現に基づく高い汎化能力を備えており,従来の記号処理の限界を打破する可能性...

A free learning portal designed to help faculty to teach, learn, and connect with others in the fields of data warehousing, DSS/BI, and database is being made available by Teradata, a division of NCR. Teradata worked closely with leading academics to create the Teradata University Network (www.teradatauniversitynetwork.com). The network is reached by faculty through a portal that includes (1) c...