PROF. WAGACHA PETER WAIGANJO

The use of ICT in agriculture within developing countries has quickly gained
popularity among development agencies, the private sector and even the government. ICT
for agriculture (ICT4Ag) services such as trade platforms, notification platforms and
advisory/extension services have been developed. This has been catalyzed by the growing
number of farmers with access to ICT devices such as mobile phones. Among the available
services, advisory/extension platforms have gained popularity among farmers an

ICT has been widely accepted and adopted as a key driver for various sectors of
the economy for both the developing and developed nations. In developing countries, there
have been multiple interventions to employ the available technology such as mobile,
wireless, radio and TV technologies in key areas that concern human development such as
health, agriculture, education and finance. The design and development practices, are
mostly borrowed from established markets with different user profiles, and do not always

Online collaborative learning provides new opportunities for student
collaboration in an online learning environment and at the same time spawns new
challenges for teachers supporting group work. With the current Course Management
Systems (CMS) such as Moodle, technology has provided online tools that include
discussions forums, chat rooms, e-mails, newsgroups, workshops, etc. These tools provide a
collaborative learning environment. To include constructivist learning in an online learning ...

Generation of three dimensional structures of macromolecules using in silico
structural modeling technologies such as homology and de novo modeling has improved
dramatically and increased the speed by which tertiary structures of organisms can be
generated. This is especially the case if a homologous crystal structure is already available.
High-resolution structures can be rapidly created using only their sequence information as
input, a process that has the potential to increase the speed of scientific discovery.

According to ITU (2012), digital divide is the difference between countries in
terms of levels of ICT development. This difference remains significant. In 2011, the ICT
Development Index (IDI) value of developed countries (6.52) was twice as high as that of
developing countries (3.24). The need to link the digital divide for universal broadband
Internet access is within the key international development goals, which include World
Summit on the Information Society (WSIS) goals and Millennium Development Goals

In many of the world's growing cities, semi-formal buses form the basis of public transit systems. However, little open and standardized data exist on these systems. The Digital Matatus project in Nairobi, Kenya set out to test whether the geo-locative capabilities of mobile technology could be used to collect data on a semi-formal transit system and whether that data could be translated into the General Transit Feed Specification (GTFS) data standard for wider use. The results of this work show that mobile technologies, particularly mobile phones, which are increasingly prevalent in developing countries, can indeed be used effectively to collect and deliver data in a modified GTFS format for semi-formal transit. Perhaps more importantly, through our work in Nairobi, we were able to identify the benefits and technical needs for developing data on semi-formal transit. Overall, the work illustrates (1) how the GTFS can be adapted to semi-formal systems and used by other cities with such transit systems, (2) that there is demand from technologists as well as transport communities for comprehensive data on semi-formal transit, (3) that releasing the data openly in the GTFS standard format can help to encourage the development of transportation applications, and (4) that including the entire transit community during the data development can create a community of users and mechanisms for institutionalizing a process of data updating and sharing. The engagement strategies our research team developed around the data collection process in Nairobi became just as important as the resulting data it produced.

In this paper we explore the impact of an intelligent grouping algorithm based on
learners' collaborative competency when compared with (a) instructor based Grade Point
Average (GPA) method level and (b) random method, on group outcomes and group
collaboration problems in an online collaborative learning environment. An intelligent
grouping algorithm has been added in a Learning Management System (LMS) which is
capable of forming heterogeneous groups based on learners' collaborative competency

Learner-centered learning theories such as active learning and collaborative
learning are highly supported by Web 2.0 technologies and they are augmenting traditional
teacher-centered approaches. New teaching pedagogies such as flipped classroom have
also embraced the use of collaborative learning where students engage in group-based
activities during class time and they embark on asynchronous video lectures after the
classroom. However, there is little research on how flipped classrooms can support online

The increased demand for higher education has made online learning popular
and appealing to many stakeholders including working staff and students. Though online
learning has gained popularity, it is still being criticized for being a faceless medium that
does little to support social interaction. Social constructivist argue that knowledge is
constructed through social activities and therefore, knowledge developed using
collaboration is more than what can be achieved by an individual alone. Online learning,

African countries are currently experiencing proliferation of mobile phone
subscriptions but no prevalence of personal computers or electricity (Parker, 2011). It is
estimated that, by the end of 2015 in Sub-Saharan Africa, the percentage of people with
mobile network access will surpass that of access to electricity in homes (Rao, 2011). This
phenomenon is also experienced in learning institutions, particularly universities, where
almost every student owns a mobile phone (Kashorda & Waema, 2009). Although there is

The essence of this chapter is to discuss theories and practices including
approaches that instructional designers consider when designing Web-based instruction for
adult learners. The importance of the chapter is to discuss best practice activities and
theories as well as technologies that enable adult online learners to be involved in the
design of their Web-based instruction. This represents recognition that adults have
accumulated a repertoire of knowledge and experiences that inevitably will enrich the

Steganography is an ancient art. It is used for security in open systems. It focuses
on hiding secret messages inside a cover medium. The most important property of a cover
medium is the amount of data that can be stored inside it without changing its noticeable
properties. There are many sophisticated techniques with which to hide, analyze, and
recover that hidden information. This research work discusses an exploration in the use of
Genetic Algorithm operators on the cover medium. We worked with text as the cover

Abstract This paper elucidates the InterlinguaPlus design and its application in bi-directional
text translations between Ekegusii and Kiswahili languages unlike the traditional translation
pairs, one-by-one. Therefore, any of the languages can be the source or target language.
The first section is an overview of the project, which is followed by a brief review of Machine
Translation. The next section discusses the implementation of the system using Carabao's
open machine translation framework and the results obtained. So far, the translation

Earlier forms of distance education were characterized by minimal social interaction
like correspondence, television, video and radio. However, the World Wide Web (WWW) and
online learning introduced the opportunity for much more social interaction, particularly
among learners, and this has been further made possible through social media in Web 2.0.
The increased availability of collaborative tools in Web 2.0 has made it possible to have
online collaborative learning realized in Higher Learning Institutions (HLIs). However,

The 8th International Workshop of the ISCA Special Interest Group on Speech and
Language Technology for Minority Languages (SALTMIL) 1 and the Fourth Workshop on
African Language Technology (AfLaT2012) 2 is held as a joint effort as part of the 2012
International Language Resources and Evaluation Conference (LREC 2012). Entitled
“Language technology for normalisation of less-resourced languages”, the workshop is
intended to continue the series of SALTMIL/LREC workshops on computational language

Research in machine translation and corpus annotation has greatly benefited from the increasing availability of word-aligned parallel corpora. This paper presents ongoing research on the development and application of the sawa corpus, a two-million-word parallel corpus English-Swahili. We describe the data collection phase and zero in on the difficulties of finding appropriate and easily accessible data for this language pair. In the data annotation phase, the corpus was semi-automatically sentence and word-aligned and morphosyntactic information was added to both the English and Swahili portion of the corpus. The annotated parallel corpus allows us to investigate two possible uses. We describe experiments with the projection of part-of-speech tagging annotation from English onto Swahili, as well as the development of a basic statistical machine translation system for this language pair, using the parallel corpus and a consolidated database of existing English-Swahili translation dictionaries. We particularly focus on the difficulties of translating English into the morphologically more complex Bantu language of Swahili.

This paper describes the development of a data-driven part-of-speech tagger and
named entity recognizer for the resource-scarce Bantu language of Kıkamba. A small
webmined corpus for Kıkamba was manually annotated for both classification tasks and
used as training material for a memory-based tagger. The encouraging experimental results
show that basic language technology tools can be developed using limit amounts of data
and state-of-the-art language-independent machine learning techniques

Abstract In this paper we present the development of a morphological analysis system for
Gikuyu Major morphological processes prevalent in Gikuyu language are explored. These
include verb inflection, verb and noun derivation and verb reduplication. In this work, finite
state transducers are used to model Gikuyu morphology. Xerox finite state tools are used to
build the lexical transducers for the system and to model rewrite rules prevalent in Gikuyu.
The system achieves an acceptable representation of Gikuyu morphology. It can correctly

Even though the Bantu language of Swahili is spoken by more than fifty million people in
East and Central Africa, it is surprisingly resource-scarce from a language technological
point of view, an unfortunate situation that holds for most, if not all languages on the
continent. The increasing amount of digitally available, vernacular data has prompted
researchers to investigate the applicability of corpus-based approaches to African language
technology. In this vein, the SAWA corpus project attempts to collect and deploy a parallel

n/a

n/a

Abstract In this paper, we describe the development of an open source spell checker for
Gıkuyu, using the Hunspell language tools. We explore the morphology of Gıkuyu and
highlight the inflection of various parts of speech, including verbs, nouns, and adjectives. In
Hunspell, surface words are realized as a set of continuation classes, with each class
providing a morpheme with a specific function. In addition, circumfixation, which is prevalent
in Gıkuyu, is implemented. Hunspell also provides for word suggestion, using character

n/a

n/a

Abstract Research in data-driven methods for Machine Translation has greatly benefited
from the increasing availability of parallel corpora. Processing the same text in two different
languages yields useful information on how words and phrases are translated from a source
language into a target language. To investigate this, a parallel corpus is typically aligned by
linking linguistic tokens in the source language to the corresponding units in the target
language. An aligned parallel corpus therefore facilitates the automatic development of a

n/a

Normal 0 false false false MicrosoftInternetExplorer4 /* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:"Times New Roman"; mso-ansi-language:#0400; mso-fareast-language:#0400; mso-bidi-language:#0400;}

HTQO VJG increasing availability of parallel corpora. Processing the same text in two
different languages yields useful information on how words and phrases are translated from
a source language into a target language. To investigate this, a parallel corpus is typically
aligned by linking linguistic tokens in the source language to the corresponding units in the
target language. An aligned parallel corpus therefore facilitates the automatic

Web-based learning systems give students the freedom to determine what to study based on
each individual learner's learning goals. These systems support learners in constructing
their own knowledge for solving problems at hand. However, in the absence of instructors,
learners often need to be supported as they learn in ways that are tailored to suit a specific
learner. Adaptive web-based learning systems fit in such situations. In order for an adaptive
learning system to be able to provide learning support, it needs to build a model of each

n/a

The orthography of many resource-scarce languages includes diacritically marked
characters. Falling outside the scope of the standard Latin encoding, these characters are
often represented in digital language resources as their unmarked equivalents. This renders
corpus compilation more difficult, as these languages typically do not have the benefit of
large electronic dictionaries to perform diacritic restoration.

This paper describes a proof-of-the-principle experiment in which maximum entropy
learning is used for the automatic induction of shallow morphological features for the
resourcescarce Bantu language of Gıkuyu. This novel approach circumvents the limitations
of typical unsupervised morphological induction methods that employ minimum-edit
distance metrics to establish morphological similarity between words. The experimental
results show that the unsupervised maximum entropy learning approach compares

n/a

n/a

n/a

n/a

Networking the development of computational resources for African languages can
be greatly advanced if researchers aim to develop tools that are to a large extent language-
independent and therefore reusable for other languages. In this paper we describe a
particular case study, namely the development of an annotated corpus of Gıkuyu, using
language-independent machine learning techniques. The general aim of our work on
Gıkuyu is two-fold: on the one hand we wish to digitally preserve this resource-scarce

Normal 0 false false false MicrosoftInternetExplorer4 /* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:"Times New Roman"; mso-ansi-language:#0400; mso-fareast-language:#0400; mso-bidi-language:#0400;}

n/a

Text-to-speech (TTS) applications have been applied in diverse areas all over the
world. Considering the fact that Swahili pronunciation is not complicated, and the language
spoken by about 45–100 million people as their first or second language,, we considered
the feasibility, and developed a Swahili Text-to-Speech (TTS) system.

Normal 0 false false false MicrosoftInternetExplorer4 /* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:"Times New Roman"; mso-ansi-language:#0400; mso-fareast-language:#0400; mso-bidi-language:#0400;}

Normal 0 false false false MicrosoftInternetExplorer4 /* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:"Times New Roman"; mso-ansi-language:#0400; mso-fareast-language:#0400; mso-bidi-language:#0400;}

In Africa, where we have enormous and varied challenges in accessing higher
education, there is need for relevant and customized content that is specific to our needs
and challenges. Most of the models that exist to address these challenges have their
limitations in terms of flexibility, time and space constraints and hence the need to address
the mitigating factors. A blend of different types of information and communication
technologies can be used in innovative ways in order to resolve some of these limitations.

Normal 0 false false false MicrosoftInternetExplorer4 /* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:"Times New Roman"; mso-ansi-language:#0400; mso-fareast-language:#0400; mso-bidi-language:#0400;}

n/a

n/a

Normal 0 false false false MicrosoftInternetExplorer4 /* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:"Times New Roman"; mso-ansi-language:#0400; mso-fareast-language:#0400; mso-bidi-language:#0400;}

The nearest neighbour classifier is one of the simplest and oldest methods for performing
general, non-parametric classification. It can be represented by the following rule: to classify
an unknown pattern, choose the class of the nearest example in the training set as
measured by a distance metric. A common extension is to choose the most common class in
the k nearest neighbours. Despite its simplicity, the nearest neighbour classifier has many
advantages over other methods as well as disadvantages.

n/a

The goal of classifier model training is not to learn an exact representation of the training
data itself, but rather to build a statistical model of the process which generates the data.
This is important if the classifier is to exhibit good generalization, that is, to make good
predictions for new inputs. Using the simple analogy of curve fitting using polynomials, we
should that a polynomial with too few coefficients gives poor predictions for new data, ie
poor generalization, since the polynomial function has too little flexibility.

n/a

The Metropolis-Hastings Markov Chain Monte Carlo algorithm uses a proposal
distribution to sample from a given target distribution. The proposal distribution has to be
tuned beforehand which is an expensive exercise. However, adaptive algorithms
automatically tune the proposal distribution based on knowledge from past samples,
reducing the tuning cost. In this study we first examine Adaptive Metropolis, Metropolis
Gaussian Adaptation and Covariance Matrix Adaptation Evolutionary Strategies.

Abstract In this paper we present the development of a morphological analysis system for
Gĩkũyũ Major morphological processes prevalent in Gĩkũyũ language are explored. These
include verb inflection, verb and noun derivation and verb reduplication. In this work, finite
state transducers are used to model Gĩkũyũ morphology. Xerox finite state tools are used to
build the lexical transducers for the system and to model rewrite rules prevalent in Gĩkũyũ.

n/a