Journal Articleshttp://repository.seku.ac.ke/xmlui/handle/123456789/22024-03-29T12:51:40Z2024-03-29T12:51:40ZMaximal activation weighted memory for aspect based sentiment analysisMokhosi, RefuoeShikali, Casper S.Qin, ZhiguangLiu, Qiaohttp://repository.seku.ac.ke/xmlui/handle/123456789/75312024-03-25T13:08:28Z2022-11-01T00:00:00ZMaximal activation weighted memory for aspect based sentiment analysis
Mokhosi, Refuoe; Shikali, Casper S.; Qin, Zhiguang; Liu, Qiao
The vast diffusion of social networks has made an unprecedented amount of user-generated data available, increasing the importance of Aspect Based Sentiment Analysis(ABSA) when extracting sentiment polarity. Although recent research efforts favor the use of self attention networks to solve the ABSA task, they still face difficulty in extracting long distance relations between non-adjacent words, especially when a sentence has more than one aspect. We propose the BERT-MAM model which approaches the ABSA task as a memory activation process regulated by memory decay and word similarity, implying that the importance of a word decays over time until it is reactivated by a similarity boost. We base experiments on the less commonly used Bidirectional Encoder Representations from Transformers (BERT), to achieve competitive results in the Laptop and Restaurant datasets.
https://doi.org/10.1016/j.csl.2022.101402
2022-11-01T00:00:00ZEnhancing African low-resource languages: Swahili data for language modellingShikali, Casper S.Mokhosi, Refuoehttp://repository.seku.ac.ke/xmlui/handle/123456789/75302024-03-25T12:57:03Z2020-08-01T00:00:00ZEnhancing African low-resource languages: Swahili data for language modelling
Shikali, Casper S.; Mokhosi, Refuoe
Language modelling using neural networks requires adequate data to guarantee quality word representation which is important for natural language processing (NLP) tasks. However, African languages, Swahili in particular, have been disadvantaged and most of them are classified as low resource languages because of inadequate data for NLP. In this article, we derive and contribute unannotated Swahili dataset, Swahili syllabic alphabet and Swahili word analogy dataset to address the need for language processing resources especially for low resource languages. Therefore, we derive the unannotated Swahili dataset by pre-processing raw Swahili data using a Python script, formulate the syllabic alphabet and develop the Swahili word analogy dataset based on an existing English dataset. We envisage that the datasets will not only support language models but also other NLP downstream tasks such as part-of-speech tagging, machine translation and sentiment analysis
https://doi.org/10.1016/j.dib.2020.105951
2020-08-01T00:00:00ZAssessment of the impact of road transport policies on air pollution and greenhouse gas emissions in KenyaMbandi, Aderiana M.Malley, Christopher S.Schwela, DietrichVallack, HarryEmberson, LisaAshmore, Mike R.http://repository.seku.ac.ke/xmlui/handle/123456789/75292024-03-25T12:26:50Z2023-01-01T00:00:00ZAssessment of the impact of road transport policies on air pollution and greenhouse gas emissions in Kenya
Mbandi, Aderiana M.; Malley, Christopher S.; Schwela, Dietrich; Vallack, Harry; Emberson, Lisa; Ashmore, Mike R.
We compile a detailed road transport inventory for greenhouse gases and air pollutants to explore energy emissions from alternative policy scenarios for the Kenya road transport sector. In 2010, road transport emissions accounted for 61% of total nitrogen oxides emissions in Kenya, 39% of fine particulate matter, 20% of carbon dioxide. In the business as usual scenario, road transport emissions increase between 4 and 31-fold from 2010 to 2050, with projected increases of motorcycles accounting for nearly all the increased pollutant emissions. Improved vehicle emission and fuel economy standards, fuel shift and investment in public transport are shown to be effective mitigation options to meet Kenya’s climate change goals with the additional benefits of better air quality and improved health.
https://doi.org/10.1016/j.esr.2023.101120
2023-01-01T00:00:00ZThe structural use of timber in construction. A reviewOluchiri, Timothy O.Sabuni, BernadetteWamalwa, C.Omondi, Bhttp://repository.seku.ac.ke/xmlui/handle/123456789/75282024-03-20T10:51:19Z2024-01-01T00:00:00ZThe structural use of timber in construction. A review
Oluchiri, Timothy O.; Sabuni, Bernadette; Wamalwa, C.; Omondi, B
Timber is one of the construction materials which when properly used, can have positive impact to both human beings and the entire ecosystem. This study therefore examines the performance, challenges and prospects of timber as a material for use in construction works. The study also highlights the sustainability benefits that are attached to the use of timber in the building system; the different types of timber used in construction; the properties of timber and the various applications of timber. From this study, it was noted that timber can be used for all types of structures if certain precautions can be observed. Some of the issues that lead to poor performance of timber in construction are; poor seasoning of wood; untreated and non-preserved wood; poor coating and technology which if can be done correctly can see timber being used for a long time. The other issues range from approval from the necessary government regulatory agencies and acceptability. Contemporary construction of tall buildings from timber, in whole or in part, suggests a growing interest in the potential for building with wood at a scale not previously seen before. As wood is the only significant building material that is grown, we have a natural inclination that building in wood is good for the environment. Building with wood does not pollute the environment. The environmental benefits of using timber are straightforward and enormous.
2024-01-01T00:00:00Z