Rabu, 14 Juli 2010
Learner’s Contribution
Readers themselves may be the best source for feedback. Teachers can check comprehension by asking their students open-ended questions, having them justify answers, or by collecting summaries. Celece-Murcia (1991) presents many approaches to analyzing text and developing understanding of the author’s intent depending on the reading. She especially advocates pre- and post- reading discussions with children to allow them to realize schemata is brought to a text and becomes modified by it.
Inessential unknown vocabulary words disrupt comprehension unless the reader recognizes that they can be ignored. Strategy is an integral part of learning, more relevant than specific linguistic knowledge. How readers solve problems is a better focus than looking at what is problematic for them (Cohen et al. 1979). Language based and schema based problems for readers are dealt with in a variety of ways researchers are only beginning to be aware of. Learners who question and monitor what they read should realize it is a natural part of good reading not a weakness of their knowledge in the new language. Block, (Ibid.) suggests recognizing the source of the problem is the first step in applying a strategy and quotes Carrell (1989) as saying that the difficulty in the application of a strategy is when it is appropriate and why this is useful.
Conclusions
Perceptions of reality are restricted by the conventions used to record them. Meaning is decoded in a mysterious process that still is not fully understood. Work done on AI has led towards a new respect for human potential by developing models of how minds work. Software attempting to imitate top-down processing alone has not resulted in perfection, lending support to other ideas. The task of reading is accomplished through an interaction of top-down and bottom-up processing. A person’s past knowledge allows text deconstruction but is simultaneously added to during the process by new information. Technology is evolving and models depicting the paradigm of gaining knowledge are being built upon (Ackley, 2001 in the Economist).
Becoming a fluent reader involves finding connections to one’s own life and making new information part of one’s own knowledge. The development of principled flexible skills that can be applied to different reading tasks is one of the most effective things from a reading class (McDonough and Shaw Ibid. p. 112). Learners as well as educators can better understand what messages are in a text by examining it with a number of approaches.
11
Schema theory offers insight on the way knowledge is constructed but is far from a complete unveiling of the mysterious process of reading.
Readers themselves may be the best source for feedback. Teachers can check comprehension by asking their students open-ended questions, having them justify answers, or by collecting summaries. Celece-Murcia (1991) presents many approaches to analyzing text and developing understanding of the author’s intent depending on the reading. She especially advocates pre- and post- reading discussions with children to allow them to realize schemata is brought to a text and becomes modified by it.
Inessential unknown vocabulary words disrupt comprehension unless the reader recognizes that they can be ignored. Strategy is an integral part of learning, more relevant than specific linguistic knowledge. How readers solve problems is a better focus than looking at what is problematic for them (Cohen et al. 1979). Language based and schema based problems for readers are dealt with in a variety of ways researchers are only beginning to be aware of. Learners who question and monitor what they read should realize it is a natural part of good reading not a weakness of their knowledge in the new language. Block, (Ibid.) suggests recognizing the source of the problem is the first step in applying a strategy and quotes Carrell (1989) as saying that the difficulty in the application of a strategy is when it is appropriate and why this is useful.
Conclusions
Perceptions of reality are restricted by the conventions used to record them. Meaning is decoded in a mysterious process that still is not fully understood. Work done on AI has led towards a new respect for human potential by developing models of how minds work. Software attempting to imitate top-down processing alone has not resulted in perfection, lending support to other ideas. The task of reading is accomplished through an interaction of top-down and bottom-up processing. A person’s past knowledge allows text deconstruction but is simultaneously added to during the process by new information. Technology is evolving and models depicting the paradigm of gaining knowledge are being built upon (Ackley, 2001 in the Economist).
Becoming a fluent reader involves finding connections to one’s own life and making new information part of one’s own knowledge. The development of principled flexible skills that can be applied to different reading tasks is one of the most effective things from a reading class (McDonough and Shaw Ibid. p. 112). Learners as well as educators can better understand what messages are in a text by examining it with a number of approaches.
11
Schema theory offers insight on the way knowledge is constructed but is far from a complete unveiling of the mysterious process of reading.
headm@mcmaster.ca
ABSTRACT
Successful implementation of e-learning is dependent on the extent to which the needs and concerns of the
stakeholder groups involved are addressed. This paper discusses e-learning, describes the needs and concerns of
the various stakeholder groups, and derives a Stakeholders' Responsiblity Matrix to summarize the
responsibilities of each stakeholder group. Fulfilling the responsibilities described in the Stakeholders’
Responsibility Matrix will address the needs and concerns of each stakeholder groups, thereby encouraging the
success of e-learning in higher education.
Keywords
E-learning, Higher education, Stakeholder analysis
Introduction
The environment of higher education is evolving. Rising costs, shrinking budgets, and an increasing need for
distance education (New Media Consortium, 2007) are causing educational institutions to reexamine the way that
education is delivered. In response to this changing environment, e-learning is being implemented more and more
frequently in higher education, creating new and exciting opportunities for both educational institutions and students.
E-learning, or electronic learning, has been defined a number of different ways in the literature. In general, e-learning
is the expression broadly used to describe “instructional content or learning experience delivered or enabled by
electronic technologies” (Ong, Lai and Wang, 2004, page 1). Some definitions of e-learning are more restrictive than
this one, for example limiting e-learning to content delivery via the Internet (Jones, 2003). The broader definition,
which will be used for the purposes of this article, can include the use of the Internet, intranets/extranets, audio- and
videotape, satellite broadcast, interactive TV, and CD-ROM, not only for content delivery, but also for interaction
among participants (Industry Canada, 2001). More recently, this definition can be further expanded to include mobile
and wireless learning applications (Kinshuk, Suhonen, Sutinen, and Goh, 2003; Lehner, Nösekabel and Lehmann,
2003).
The e-learning models of higher education today find their roots in conventional distance education. Initially
introduced to allow individuals in remote and rural areas to gain access to higher education, distance learning has
evolved significantly over time. Technological advancement has been the major inspiration for change, beginning
with the integration of radio broadcasting in the 1920’s (Huynh, Umesh and Valachich, 2003). More recently, the
advent of the Internet has enabled tremendous innovation in the delivery of post secondary education (Gunasekaran,
McNeil and Shaul, 2002; Teo and Gay, 2006). As time goes by, more and more people gain access to the Internet,
the cost of computer ownership decreases, and overall computer literacy increases (Huynh et al., 2003). These trends
provide educational institutions an ideal channel for the delivery of educational content.
Dimensions of E-Learning
The extent of e-learning technology use in course delivery varies widely. The variations in the configuration of elearning
offerings can be described through a number of attributes, as listed in Table 1 below. These attributes can be
classified into the dimensions of synchronicity, location, independence, and mode. An e-learning course component
can be described by indicating which one of the two attribute values from each dimension is applicable.
E-learning can be synchronous (real-time) or asynchronous (flex-time). Synchronous e-learning includes technology
such as video conferencing and electronic white boards (Romiszowski, 2004), requiring students to be present at the
ABSTRACT
Successful implementation of e-learning is dependent on the extent to which the needs and concerns of the
stakeholder groups involved are addressed. This paper discusses e-learning, describes the needs and concerns of
the various stakeholder groups, and derives a Stakeholders' Responsiblity Matrix to summarize the
responsibilities of each stakeholder group. Fulfilling the responsibilities described in the Stakeholders’
Responsibility Matrix will address the needs and concerns of each stakeholder groups, thereby encouraging the
success of e-learning in higher education.
Keywords
E-learning, Higher education, Stakeholder analysis
Introduction
The environment of higher education is evolving. Rising costs, shrinking budgets, and an increasing need for
distance education (New Media Consortium, 2007) are causing educational institutions to reexamine the way that
education is delivered. In response to this changing environment, e-learning is being implemented more and more
frequently in higher education, creating new and exciting opportunities for both educational institutions and students.
E-learning, or electronic learning, has been defined a number of different ways in the literature. In general, e-learning
is the expression broadly used to describe “instructional content or learning experience delivered or enabled by
electronic technologies” (Ong, Lai and Wang, 2004, page 1). Some definitions of e-learning are more restrictive than
this one, for example limiting e-learning to content delivery via the Internet (Jones, 2003). The broader definition,
which will be used for the purposes of this article, can include the use of the Internet, intranets/extranets, audio- and
videotape, satellite broadcast, interactive TV, and CD-ROM, not only for content delivery, but also for interaction
among participants (Industry Canada, 2001). More recently, this definition can be further expanded to include mobile
and wireless learning applications (Kinshuk, Suhonen, Sutinen, and Goh, 2003; Lehner, Nösekabel and Lehmann,
2003).
The e-learning models of higher education today find their roots in conventional distance education. Initially
introduced to allow individuals in remote and rural areas to gain access to higher education, distance learning has
evolved significantly over time. Technological advancement has been the major inspiration for change, beginning
with the integration of radio broadcasting in the 1920’s (Huynh, Umesh and Valachich, 2003). More recently, the
advent of the Internet has enabled tremendous innovation in the delivery of post secondary education (Gunasekaran,
McNeil and Shaul, 2002; Teo and Gay, 2006). As time goes by, more and more people gain access to the Internet,
the cost of computer ownership decreases, and overall computer literacy increases (Huynh et al., 2003). These trends
provide educational institutions an ideal channel for the delivery of educational content.
Dimensions of E-Learning
The extent of e-learning technology use in course delivery varies widely. The variations in the configuration of elearning
offerings can be described through a number of attributes, as listed in Table 1 below. These attributes can be
classified into the dimensions of synchronicity, location, independence, and mode. An e-learning course component
can be described by indicating which one of the two attribute values from each dimension is applicable.
E-learning can be synchronous (real-time) or asynchronous (flex-time). Synchronous e-learning includes technology
such as video conferencing and electronic white boards (Romiszowski, 2004), requiring students to be present at the
Proceedings of the Workshop on Statistical Machine Translation, pages 47–54,
New York City, June 2006. c
2006 Association for Computational Linguistics
Searching for alignments in SMT. A novel approach based on an Estimation
of Distribution Algorithm ¤
Luis Rodr´ıguez, Ismael Garc´ıa-Varea, Jos´e A. G´amez
Departamento de Sistemas Inform´aticos
Universidad de Castilla-La Mancha
luisr@dsi.uclm.es, ivarea@dsi.uclm.es, jgamez@dsi.uclm.es
Abstract
In statistical machine translation, an alignment
defines a mapping between the
words in the source and in the target sentence.
Alignments are used, on the one
hand, to train the statistical models and, on
the other, during the decoding process to
link the words in the source sentence to the
words in the partial hypotheses generated.
In both cases, the quality of the alignments
is crucial for the success of the translation
process. In this paper, we propose an algorithm
based on an Estimation of Distribution
Algorithm for computing alignments
between two sentences in a parallel
corpus. This algorithm has been tested
on different tasks involving different pair
of languages. In the different experiments
presented here for the two word-alignment
shared tasks proposed in the HLT-NAACL
2003 and in the ACL 2005, the EDAbased
algorithm outperforms the best participant
systems.
1 Introduction
Nowadays, statistical approach to machine translation
constitutes one of the most promising approaches
in this field. The rationale behind this approximation
is to learn a statistical model from a parallel
corpus. A parallel corpus can be defined as a set
¤This work has been supported by the Spanish Projects
JCCM (PBI-05-022) and HERMES 05/06 (Vic. Inv. UCLM)
of sentence pairs, each pair containing a sentence in
a source language and a translation of this sentence
in a target language. Word alignments are necessary
to link the words in the source and in the target
sentence. Statistical models for machine translation
heavily depend on the concept of alignment,
specifically, the well known IBM word based models
(Brown et al., 1993). As a result of this, different
task on aligments in statistical machine translation
have been proposed in the last few years (HLTNAACL
2003 (Mihalcea and Pedersen, 2003) and
ACL 2005 (Joel Martin, 2005)).
In this paper, we propose a novel approach to deal
with alignments. Specifically, we address the problem
of searching for the best word alignment between
a source and a target sentence. As there is
no efficient exact method to compute the optimal
alignment (known as Viterbi alignment) in most of
the cases (specifically in the IBM models 3,4 and 5),
in this work we propose the use of a recently appeared
meta-heuristic family of algorithms, Estimation
of Distribution Algorithms (EDAs). Clearly, by
using a heuristic-based method we cannot guarantee
the achievement of the optimal alignment. Nonetheless,
we expect that the global search carried out
by our algorithm will produce high quality results
in most cases, since previous experiments with this
technique (Larra˜naga and Lozano, 2001) in different
optimization task have demonstrated. In addition to
this, the results presented in section 5 support the
approximation presented here.
This paper is structured as follows. Firstly, Statistical
word alignments are described in section 2.
Estimation of Distribution Algorithms (EDAs) are
47
New York City, June 2006. c
2006 Association for Computational Linguistics
Searching for alignments in SMT. A novel approach based on an Estimation
of Distribution Algorithm ¤
Luis Rodr´ıguez, Ismael Garc´ıa-Varea, Jos´e A. G´amez
Departamento de Sistemas Inform´aticos
Universidad de Castilla-La Mancha
luisr@dsi.uclm.es, ivarea@dsi.uclm.es, jgamez@dsi.uclm.es
Abstract
In statistical machine translation, an alignment
defines a mapping between the
words in the source and in the target sentence.
Alignments are used, on the one
hand, to train the statistical models and, on
the other, during the decoding process to
link the words in the source sentence to the
words in the partial hypotheses generated.
In both cases, the quality of the alignments
is crucial for the success of the translation
process. In this paper, we propose an algorithm
based on an Estimation of Distribution
Algorithm for computing alignments
between two sentences in a parallel
corpus. This algorithm has been tested
on different tasks involving different pair
of languages. In the different experiments
presented here for the two word-alignment
shared tasks proposed in the HLT-NAACL
2003 and in the ACL 2005, the EDAbased
algorithm outperforms the best participant
systems.
1 Introduction
Nowadays, statistical approach to machine translation
constitutes one of the most promising approaches
in this field. The rationale behind this approximation
is to learn a statistical model from a parallel
corpus. A parallel corpus can be defined as a set
¤This work has been supported by the Spanish Projects
JCCM (PBI-05-022) and HERMES 05/06 (Vic. Inv. UCLM)
of sentence pairs, each pair containing a sentence in
a source language and a translation of this sentence
in a target language. Word alignments are necessary
to link the words in the source and in the target
sentence. Statistical models for machine translation
heavily depend on the concept of alignment,
specifically, the well known IBM word based models
(Brown et al., 1993). As a result of this, different
task on aligments in statistical machine translation
have been proposed in the last few years (HLTNAACL
2003 (Mihalcea and Pedersen, 2003) and
ACL 2005 (Joel Martin, 2005)).
In this paper, we propose a novel approach to deal
with alignments. Specifically, we address the problem
of searching for the best word alignment between
a source and a target sentence. As there is
no efficient exact method to compute the optimal
alignment (known as Viterbi alignment) in most of
the cases (specifically in the IBM models 3,4 and 5),
in this work we propose the use of a recently appeared
meta-heuristic family of algorithms, Estimation
of Distribution Algorithms (EDAs). Clearly, by
using a heuristic-based method we cannot guarantee
the achievement of the optimal alignment. Nonetheless,
we expect that the global search carried out
by our algorithm will produce high quality results
in most cases, since previous experiments with this
technique (Larra˜naga and Lozano, 2001) in different
optimization task have demonstrated. In addition to
this, the results presented in section 5 support the
approximation presented here.
This paper is structured as follows. Firstly, Statistical
word alignments are described in section 2.
Estimation of Distribution Algorithms (EDAs) are
47
Langganan:
Postingan (Atom)