65 research outputs found
Influences of artificial light on mating of black soldier fly (Hermetia illucens)—a review
No full textBlack soldier fly (Hermetia illucens) is a potential insect species which can convert biodegradable materials and some indigestible organic waste into valuable biomass. Because of having good quality of fat and protein, its production and use in animal feed are being extended day by day. To fulfill the future demand re-searchers are trying to find out the successful mass rearing techniques of H. illucens in laboratory or indoor condition. However, the most critical part of H. illucens mass production is obtaining successful mating. This insect is very sensitive to light. It prefers direct sunlight for its successful mating however, artificial light has substantial effects on its mating behaviors. It was reported that light quality, intensity, duration have signifi-cant influences on the H. illucens successful mating and fertilized egg production. This review brings in forth all the information about artificial light effects on H. illucens adults for their successful mating towards the mass production in indoor condition.The study was supported by Sylhet Agricultural University Research System (SAURES) and the Department of Entomology, Sylhet Agricultural University, Bangladesh.Awal, Md. Rabiul; Rahman, Md Masudur; Choudhury, Md Abdur; Hasan, Md Mehedi; Rahman, Towfiq; Mondal, Md Fuad. (2022). Influences of artificial light on mating of black soldier fly (Hermetia illucens)—a review. Retrieved from the University Digital Conservancy, 10.1007/s42690-022-00786-7
Tackling Social Value Tasks with Multilingual NLP
No full textIn recent years, deep learning applications have shown promise in tackling social value tasks such as hate speech and misinformation in social media. Neural networks provide an efficient automated solution that has replaced hand-engineered systems. Existing studies that have explored building resources, e.g. datasets, models, and NLP solutions, have yielded significant performance. However, most of these systems are limited to providing solutions in only English, neglecting the bulk of hateful and misinformation content that is generated in other languages, particularly so-called low-resource languages that have a low amount of labeled or unlabeled language data for training machine learning models (e.g. Turkish). This limitation is due to the lack of a large collection of labeled or unlabeled corpora or manually crafted linguistic resources sufficient for building NLP systems in these languages.
In this thesis, we set out to explore solutions for low-resource languages to mitigate the language gap in NLP systems for social value tasks. This thesis studies two tasks. First, we show that developing an automated classifier that captures hate speech and nuances in a low-resource language variety with limited data is extremely challenging. To tackle this, we propose HateMAML, a model-agnostic meta-learning-based framework that effectively performs hate speech detection in low resource languages. The proposed method uses a self-supervision strategy to overcome the limitation of data scarcity and produces a better pre-trained model for fast adaptation to an unseen target language. Second, this thesis aims to address the research gaps in rumour detection by proposing a modification over the standard Transformer and building on a multilingual pre-trained language model to perform rumour detection in multiple languages. Specifically, our proposed model MUSCAT prioritizes the source claims in multilingual conversation threads with co-attention transformers. Both of these methods can be seen as the incorporation of efficient transfer learning methods to mitigate issues in model training with small data.
The findings yield accurate and efficient transfer learning models for low-resource languages. The results show that our proposed approaches outperform the state-of-the-art baselines in the cross-domain multilingual transfer setting. We also conduct ablation studies to analyze the characteristics of proposed solutions and provided empirical analysis outlining the challenges of data collection to performing detection tasks in multiple languages
Recommended from our members
Cross-equatorial heat transports and the monsoons
No full textMonsoons, the lifeline for billions of people, result from the cross-equatorial Hadley
circulation (HC), which imports moisture into monsoon regions and exports heat. The inter-annual relationships between cross-equatorial heat transports and monsoons are largely unexplored, which is vital for water resources and climate stability.
Using reanalysis data, it is revealed that increased cross-equatorial atmospheric heat
transport (CE-AHT) during JJA weakens the Indian, West African, and North American
monsoons by contracting the HC, shifting the subtropical jets and Inter-tropical Convergence Zone (ITCZ) equatorward, weakening the Walker circulation, and inducing El Nino- ˜
like conditions. During DJF, the Australian, South African, and South American monsoons
strengthen as the HC expands, the ITCZ and subtropical jets move poleward, strengthening
the Walker circulation, and promoting La Nina-like conditions. Global monsoon precipi- ˜
tation increases in both JJA and DJF. Conversely, increased cross-equatorial oceanic heat
transport (CE-OHT) in JJA and DJF elicits effects opposite to those of CE-AHT.
A few CMIP6 models simulate the historical trends observed in reanalysis for CE-AHT
and CE-OHT. However, the relationship between CE-AHT, CE-OHT, and monsoons in most
models mirrors reanalysis patterns.
Under the shared socioeconomic pathways 5-8.5 scenario, most CMIP6 models exhibit
upward CE-AHT and downward CE-OHT trends in JJA, but these trends are downward
and insignificant for DJF. In JJA, all models indicate increases in Indian monsoon precipitation for 2080–2100 compared to 1995–2014, while disagreement arises regarding the West
African and North American monsoons. In multi-model space, greater CE-AHT weakens
Indian monsoon rainfall, while greater CE-OHT weakens global monsoon rainfall. In DJF,
models differ on Australian, South African, and South American monsoon precipitation
changes.
This research suggests that changes in CE-AHT and CE-OHT are deeply intertwined with
monsoons, and future warming could alter monsoon dynamics and heighten uncertainty in
climate modelling and predictions
Perceived Detrimental Factors Affecting Undergraduate Accounting Students’ Academic Performance
No full textThough good academic performance is a common expectation from the students, guardians, and educational administrations, many students may not achieve good academic results. Several detrimental factors act as barriers to students’ good academic performance. Therefore, it is the concern of educational institutions to detect the detrimental factors and find out effective solutions. This study explores the undergraduate accounting students’ perceptions of the detrimental factors that negatively affect the academic performance of the students at Moulvibazar Government Women's College (MGWC), Bangladesh. A total of 45 students from undergraduate accounting 1st year through 3rd year participated in the questionnaire survey of this study and expressed opinions regarding the detrimental factors to their academic performance. The study found that several student factors (e.g., frequent absenteeism, lack of motivation, disliking to course), family factors (e.g., parental academic background, family economic status), college factors (e.g., over enrolment, inadequate campus accommodation, poor sanitation, poor library facilities) and teacher factors (e.g., lack of teachers, lack of seriousness among teachers) significantly affects students’ academic performance. This study recommends that the college authority should pay more attention to ensuring sufficient college facilities and the government should provide more funds for increasing students’ accommodation facilities, transportation facilities, building the required infrastructure, and recruiting sufficient teachers for the department of accounting at MGWC. Keywords: Detrimental Factors, Academic Performance, Undergraduate, Accounting Students. DOI: 10.7176/JEP/13-11-09 Publication date: April 30th 202
Region identification based multidimensional optimization of the performance of semiconductor laser amplifier
No full textNetwork-on-Chip implementation of Midimew-Connected Mesh Network
No full textArchitecture of interconnection network plays a
significant role in the performance and energy consumption of
Network-on-Chip (NoC) systems. In this paper we propose
NoC implementation of Midimew-connected Mesh Network
(MMN). MMN is a Minimal Distance Mesh with Wrap-around
(Midimew) links network of multiple basic modules, in which
the basic modules are 2D-mesh networks that are
hierarchically interconnected for higher-level networks. For
implementing all the links of level-3 MMN, minimum 4 layers
are needed which is feasible with current and future VLSI
technologies. With innovative combination of diagonal and
hierarchical structure, MMN possesses several attractive
features including constant node degree, small diameter, low
cost, small average distance, and moderate bisection width
than that of other conventional and hierarchical
interconnection networks
Midimew-connected mesh network
No full textMidimew-Connected Mesh Network (MMN) is a MInimal DIstance MEsh with Wrap-around links (MIDIMEW) network of multiple basic modules, in which the basic modules are 2D-mesh networks that are hierarchically interconnected for higher-level networks. In this poster, we present the structure and Network-on-Chip (NoC) implementation of the MMN. It is shown that the proposed MMN with innovative combination of diagonal and hierarchical structure, possesses several attractive features, including constant degree, small diameter, small average distance, moderate bisection width than that of other conventional interconnection networks and requires less amount of wires for the implementation of the physical links. For the implementation of NoC of MMN at least four layers are needed to implement all links of MMN level-3 network which is feasible with current and future VLSI technologies
First record of important biological parameters of Badis badis: A small indigenous species in Bangladesh
No full textA total of 286 Badis badis were collected from the Sutiyahali Reservoir in Mymensingh from January to December 2022, and their sex ratios, first sexual maturity, length-weight relationships and condition factors were evaluated. The weight and length of B. badis varied from 0.81 to 1.01g (0.89±0.30) and 4.08 to 4.60cm (4.36±0.31), respectively. Logistic curves depicting a sex ratio and 50% maturity (L50) estimated at 4.5cm for females and 4.05cm for males, as well as males reaching first sexual maturity with a shorter length than females. Regression coefficients in every month differ significantly (p<0.05), according to the regression equations. Each month, the values of the exponent b were less than 3 (b<3), with the highest value of b recorded in August (2.80) and the lowest value recorded in January (2.33). This led to a monthly negative allometric growth being seen. A strong positive relationship is evident from the coefficient of determination (r2) values, which ranged from 0.92-0.98 with an average of 0.961. During the study, the average condition factor (Kn) value for B. badis was found to be 1.02±0.13, which is a positive indicator of the fish's physical well-being. The condition factor values varied between 0.84 to 1.39, making it abundantly clear that B. badis are in good health and the waterbody is an ideal habitat for their survival. Relative condition factor (Kr) values, which varied between studies and ranged from 0.78 to 1.01, also exhibited a noteworthy difference (p<0.05). For its long-term management, the above findings will be very helpful
EEG Channel Correlation Based Model for Emotion Recognition
No full textEmotion recognition using Artificial Intelligence (AI) is a fundamental prerequisite to improve Human-Computer Interaction (HCI). Recognizing emotion from Electroencephalogram (EEG) has been globally accepted in many applications such as intelligent thinking, decision-making, social communication, feeling detection, affective computing, etc. Nevertheless, due to having too low amplitude variation related to time on EEG signal, the proper recognition of emotion from this signal has become too challenging. Usually, considerable effort is required to identify the proper feature or feature set for an effective feature-based emotion recognition system. To extenuate the manual human effort of feature extraction, we proposed a deep machine-learning-based model with Convolutional Neural Network (CNN). At first, the one-dimensional EEG data were converted to Pearson's Correlation Coefficient (PCC) featured images of channel correlation of EEG sub-bands. Then the images were fed into the CNN model to recognize emotion. Two protocols were conducted, namely, protocol-1 to identify two levels and protocol-2 to recognize three levels of valence and arousal that demonstrate emotion. We investigated that only the upper triangular portion of the PCC featured images reduced the computational complexity and size of memory without hampering the model accuracy. The maximum accuracy of 78.22% on valence and 74.92% on arousal were obtained using the internationally authorized DEAP dataset.Full Tex
Development of self-charging unmanned aerial vehicle system using inductive approach
No full textThis paper presents an alternative approach to power up unmanned aerial vehicle (UAV) system using inductive approach. The main issue of utilizing UAV in any application especially in precision agriculture is the lifetime of the battery. This limits the flight time of the UAV which makes the system is unable to be efficiently applied for precision agriculture purpose. Hence, this paper proposes a new approach of powering UAV system by using so called inductive power transfer (IPT) technology. Through this approach, the system can be powered up wirelessly with no physical link in between transmitter and receiver. To be specific, class E inverter circuit has been designed together with impedance matching circuit to ensure higher efficiency is obtained. Finally, a prototype of IPT system for powering up the UAV system was successfully developed, which is able to transmit 23.32 W of power at 1 MHz operating frequency from 12 V input supply. The system achieved up to 95.73% efficiency
- …
