AI Service Research Articles

EVALUATION OF REPRODUCTIVE PERFORMANCE AND MILK PRODUCTION POTENTIALS OF LOCAL COWS UNDER FARMER’S MANAGEMENT IN AND AROUND JINKA TOWN IN SOUTH OMO ZONE, SOUTHERN ETHIOPIA

The study was conducted in and around Jinka Town to investigate the reproductive and productive performances of local cows under the farmer’s management. Six kebeles were selected from in and around Jinka town by simple random sampling technique. One-hundred and twenty households (60 from each) were used for the household survey. Data were collected through semi-structured questionnaires, focus group discussions, and secondary sources. The data were analyzed by using the SPSS version 20 and the mean comparison was made by LSD. An average age at first mating for male cattle was significantly (p<0.05) higher in Jinka town than around it. Average number of calves produced in lifetime was significantly higher around kebeles than in Jinka town. The overall mean±SE age at sexual maturity for both female and male cattle was 3.75±.05 and 3.38±.05 year, respectively. The overall mean±SE milk production per cow per was 1.28±0.35 liter per a day. Lactation milk yield per cow was significantly higher in town than around kebeles of Jinka town. During focus discussion and interview, the most breeding practices problems that hinder herd productivity were long calving interval, delayed age at first calving, postpartum anoestrous, seasonal fluctuation, the genetic makeup of animals, inadequate AI services and lack of skilled manpower. Generally, milk production in and around Jinka town is of subsistence type. On station investigation of production and reproduction performances of the local cattle is recommended through correlation studies and regular monitoring of the population to select the best animals, pasture improvement, and introducting technology that will enhance nutritive value of crop residue and dry season feed resources are needed.

Estimation of Genetic Parameters for Heifer and Cow Fertility Traits Derived from On-Farm AI Service Records of South African Holstein Cattle.

Simple SummaryFemale fertility has gained significant attention and is incorporated in dairy cattle breeding objectives worldwide. Currently, age at first calving (AFC) and calving interval (CI) are the only traits used as indicators of fertility in the genetic evaluations of South African dairy cattle, but these traits are greatly affected by the breeder’s decisions and calving interval is available late in an animal’s life which delays selection decisions. This study investigated the possible use of additional fertility traits derived from artificial insemination records. The fertility traits examined in this study could be used in addition to age at first calving and calving interval to select for improved fertility in dairy cattle.This study aimed to derive additional fertility traits from service data and estimate their genetic parameters for their possible inclusion in the South African Holstein cattle breeding programs. Service records (n = 64,464) were collected from 18 South African Holstein herds using on-farm automated milk recording systems. Using a multivariate model, the data were used to estimate heritabilities and correlations among several fertility traits. The pedigree data consisted of information on 18,592 animals born between 1981 and 2013. Heritability estimates observed were low to moderate (0.02 ± 0.00 to 0.24 ± 0.00), indicating that there is some genetic basis for the explored fertility traits to warrant selection. The genetic correlations observed between fertility traits were generally favorable, with some high correlations between age at first service (AFS) and services per conception for heifers (SPCh) (0.73 ± 0.00) and between days from calving to first service (CFS) and services per conception for cows (SPC) (0.90 ± 0.01). Positive genetic correlations indicate that improvement in one trait is coupled with a correlated genetic increase in another trait. The studied fertility traits could be used in addition to AFC and CI to serve as a basis for the selection of reproduction in dairy cattle to minimize selection bias.

AI Service and Emotion

AI in service can be for routine mechanical tasks, analytical thinking tasks, or empathetic feeling tasks. We provide a conceptual framework for the customer, firm, and interactional use of AI for empathetic tasks at the micro-, meso-, and macro-levels. Emotions resulting from AI service interactions can include basic emotions (e.g., joy, sadness, and fear), self-conscious emotions (e.g., pride, guilt, embarrassment), and moral emotions (e.g., contempt, righteous anger, social disgust). These emotions are mostly likely to occur during frontline interactions in which both firms and customers use AI, a phenomenon called “AI as customer.” The analysis level of AI service and emotion can be at the macro-level in which AI is transforming the service economy into a feeling economy, at the meso-level in which firms can use “thoughtful AI” to make the employees’ and customers’ lives a little bit better by brightening their days, and at the micro-level in which customers can experience basic, self-conscious, and moral emotions from interactions with service AI.

KYC Principles Embedded in AI-enabled Mobile Banking in the Face of Money-Laundering

Anti-money laundering and financial terrorism laws create a major dilemma for banks, where the latter could be held liable as constructive trustees for the rightful owner of the corrupted funds. This paper aims to find out ways to maintain compliance to the Know Your Customer (KYC) principles via an Artificial Intelligence AI-enabled mobile banking service. The intervention research was conducted in the Middle East precisely in a private-owned bank. The results show that the bank’s compliance program can benefit from the AI banking services as enablers to fight Money-laundering. For this reason, the intervention research management consulting project comprises of valuing the impact of implementing a Socio-economic Approach to Management to synchronize the human potential and AI banking services to sustain the compliance program (FATF, 2012). This anticipated improvement in performance would increase the AI services and KYC implementation return on investment thus ensuring compliance and competitive advantage.

Cloud AI: A comparison of specimen image data extraction processes

The Natural History Museum (NHM) of London has embarked on an ambitious programme to digitise the 80 million specimens in its collection, releasing them through the NHM data portal and the global biodiversity research community. As part of the digitisation process, data is transcribed from specimen labels to capture the vital taxonomic and collection event data. Accurate human transcription is slow and the NHM, like many institutions, has been exploring machine learning (ML) for automated specimen analysis and label data capture. This process requires many different models, chained in series: semantic segmentation to identify specimen and label regions of interest; optical character recognition to identify text on labels; natural language processing to extract entities from the text. As part of SYNTHESYS+, the NHM has been building the Specimen Data Refinery (SDR) (Smith et al. 2019) - a workflow engine for chaining ML models, each performing one atomic task in the data extraction process. The SDR is now in public beta, and we present evaluation metrics from our initial testing. Alongside the SDR project, the NHM has been exploring cloud-based artificial intelligence tools for specimen digitisation, using Google and Amazon technologies. We present an analysis of these different approaches, comparing the results from third-party AI services with models developed specifically for the biodiversity and natural history collection domains. With large corporates providing comparatively low-cost access to AI compute resources and models transferrable to many specimen image digitisation tasks, is developing bespoke solutions still required?

ApproxIFER: A Model-Agnostic Approach to Resilient and Robust Prediction Serving Systems

Due to the surge of cloud-assisted AI services, the problem of designing resilient prediction serving systems that can effectively cope with stragglers and minimize response delays has attracted much interest. The common approach for tackling this problem is replication which assigns the same prediction task to multiple workers. This approach, however, is inefficient and incurs significant resource overheads. Hence, a learning-based approach known as parity model (ParM) has been recently proposed which learns models that can generate ``parities’’ for a group of predictions to reconstruct the predictions of the slow/failed workers. While this learning-based approach is more resource-efficient than replication, it is tailored to the specific model hosted by the cloud and is particularly suitable for a small number of queries (typically less than four) and tolerating very few stragglers (mostly one). Moreover, ParM does not handle Byzantine adversarial workers. We propose a different approach, named Approximate Coded Inference (ApproxIFER), that does not require training any parity models, hence it is agnostic to the model hosted by the cloud and can be readily applied to different data domains and model architectures. Compared with earlier works, ApproxIFER can handle a general number of stragglers and scales significantly better with the number of queries. Furthermore, ApproxIFER is robust against Byzantine workers. Our extensive experiments on a large number of datasets and model architectures show significant degraded mode accuracy improvement by up to 58% over ParM.

AI Snap! Blocks for Speech Input and Output, Computer Vision, Word Embeddings, and Neural Net Creation, Training, and Use

We will demonstrate blocks integrated into Snap! capable of a wide range of AI services, interactive AI programming guides, and a selection from thirty sample projects. Sessions and workshops in both school settings and informal learning contexts have been held in many countries. The full version of this paper includes descriptions of the Snap! blocks and unpublished descriptions of student experiences in India.

When do AI chatbots lead to higher customer satisfaction than human frontline employees in online shopping assistance? Considering product attribute type

The increasing adoption of AI chatbots in online shopping assistance, as a complement or substitute for human frontline employees (HFLEs), leads to the question whether HFLEs perform better than AI service robots and why. From the perspective of product attribute type (experiential/functional) and focusing on customer satisfaction, this study explores how the impact of service agent on customer satisfaction varies along with product attribute type. A scenario-based experiment was designed and completed by 567 participants. Although HFLEs lead to higher customer satisfaction when the product attribute is experiential, AI chatbots perform better than HFLEs when the product attribute is functional. We make use of perceived information quality, perceived waiting time, and positive emotions, three determinants of customer satisfaction, to explain the variation of the role of different service agent types. The findings offer useful implications for companies when selecting service agent types in online shopping assistance.

A Study on Major Service Items of Consumers and Companies Using Convergence Technology in the Intelligent Age

The purpose of this study is to investigate and analyze quality evaluation in the age of artificial intelligence services, and to present a plan. Preliminary quality evaluation items were derived through research on opinions of experts by researching various documents related to artificial intelligence and analyzing cases. Then, related experts were selected and research was conducted using the Delphi technique. Through this, service quality evaluation items suitable and appropriate for the age of artificial intelligence were derived and proposed.Using AI service quality evaluation items, consumers and companies can pursue very beneficial convenience and benefits. This can increase customer satisfaction and increase corporate sales or profits

The Role of Higher Education for the Ethical AI Society

AI is creating numerous ethical issues. In the upcoming AI society, we will have to deal with the Influence of AI correctly. To this end, ethical education is needed for experts who learn and develop AI technology. Above all, this is because their ethical capabilities lead to ethical AI technologies and services. Therefore, this study sought the role and direction of AI ethics education in the field of higher education that educates AI experts.

Understanding customer's meaningful engagement with AI-powered service robots

PurposeThe purpose of this study is to empirically investigate what aspects of service robot interactions with customers can lead to meaningful outcomes in the view of customers. The study examines functional and emotional elements of AI service robots in terms of meaningful outcomes.Design/methodology/approachThis study highlights AI service robots' meaningful outcomes as a viable research problem and proposes a research model utilizing the Stimulus-Organism-Response (SOR) framework. As an empirical approach, 260 datasets were collected from customers who have experience with AI service restaurants in China.FindingsThe study examines the functional and emotional elements of AI-powered service robots on the attitude of and meaningful outcomes for customers. The results showed that the emotional (perceived friendliness and perceived coolness) and functional (perceived safety and robot competence) attributes of human–robot interactions (HRI) significantly affect the attitude toward using service robots. Second, the attitude toward using service robots significantly influences the experiential outcome and instrumental outcome of meaningful engagement.Research limitations/implicationsThis study highlights two elements (i.e. functional and emotional) of HRI effectiveness using two metrics: experiential and performance outcomes. Future studies should generalize the research findings of service robots in the current study using a larger quantity of data from various service fields.Originality/valueAs the first empirical study highlighting the customer experience with service robots, this study opens up a feasible research direction for the service industry to pursue in terms of conducting HRI studies from the view of customers. It identifies a research model pursuant to customers' experience with HRI in creating meaningful outcomes and it theoretically extends the SOR model to the hospitality study, focusing on the HRI issue.

Unraveling the Hidden Environmental Impacts of AI Solutions for Environment Life Cycle Assessment of AI Solutions

In the past ten years, artificial intelligence has encountered such dramatic progress that it is now seen as a tool of choice to solve environmental issues and, in the first place, greenhouse gas emissions (GHG). At the same time, the deep learning community began to realize that training models with more and more parameters require a lot of energy and, as a consequence, GHG emissions. To our knowledge, questioning the complete net environmental impacts of AI solutions for the environment (AI for Green) and not only GHG, has never been addressed directly. In this article, we propose to study the possible negative impacts of AI for Green. First, we review the different types of AI impacts; then, we present the different methodologies used to assess those impacts and show how to apply life cycle assessment to AI services. Finally, we discuss how to assess the environmental usefulness of a general AI service and point out the limitations of existing work in AI for Green.

179 Sire Distribution of Calves in a Herd with use of Fixed Time Artificial Insemination Followed by Immediate Bull Exposure for Natural Service

Abstract Use of fixed time AI followed by immediate exposure of females to bulls for natural service can be a useful management strategy for commercial cow-calf producers to limit labor and time related to bull turnout as well as increase pregnancy rates earlier in the breeding season. Considering potential influence of factors such as bull fertility and time to and length of estrus in females, expectations for outcomes in natural service sire versus AI sire parentage is relatively unknown. We hypothesized that of calves born in the first 14 d of the calving season, less than 5% would be sired by natural service bulls exposed to females immediately following fixed time AI. Heifers (n = 191) and cows (n = 156) were synchronized and inseminated using the 7-d CO-Synch + CIDR fixed time AI protocol. All females were inseminated by the same AI technician using a single sire for heifers and a different single sire for cows. All females were exposed to bulls immediately following insemination. DNA was collected from a random subset of calves (Calves born from heifers n = 59; Calves born from cows n = 89) born in the first 14 d of the calving season for parentage analysis to estimate the overall percentage of calves sired by AI and natural service. Chi-Square tests were used to determine if the percentage of calves actually sired by natural service differed from the hypothesized 5% or less. Among calves born from heifers, the actual percentage sired by natural service (5.1%; n = 3/59) was similar (P = 0.98) to our hypothesized percentage. Among calves born from cows, the actual percentage sired by natural service (14.6%; n = 13/89) was greater (P < 0.001) than our hypothesized percentage of 5%. If commercial producers use fixed time AI followed by immediate bull exposure in cows, natural service bulls may sire more calves early in the calving season than expected.

Assessment of Preferable Breeding Method Use and Related Limitations under Smallholder Dairy Farmers’ Conditions: A Case Study in Selected Districts Southern Highland Zone of Tanzania

A study was carried out to assess status and preference of smallholder dairy farmers on breeding method use under smallholder dairy cattle production in Southern Highland zone of Tanzania. Three districts namely Njombe, Mbeya and Mbozi were purposively selected for data collection. Data were collected using a questionnaire. A total of 180 dairy farmers in three districts were voluntary interviewed. Results showed that many of dairy farmers’ age range between 41 to 50 years (33%) of respondents, 48% of respondents (48.9%) had primary education, many farmers were found not to be working in groups (82.2%) and majority had never received any training on dairy husbandry (81.1%).
 Zero grazing was the main management system (97.2%) used by farmers in all districts. Cows’ sheds were partially roofed, and the floors were soiled (48.9%). Mean dairy herd size was 5.21 per household and natural service was the predominant breeding method (55%) under smallholder farms production system. Farmers’ main reason for using AI was to improve genetic potential of their animal (48.8%). Under this system the major challenges of AI service were inefficient semen supply and poor heat detection. Natural services are widely used since it is easily accessible and cheaper. It therefore concluded that use of natural service is still a predominant method of choice in breeding due to unreliability and inefficiency associated with AI service in the Southern Highland zone of Tanzania.

The role of the human-robot interaction in consumers’ acceptance of humanoid retail service robots

The Retail Service Robot (RSR) is a humanoid robot that uses AI service automation to provide customized shopping assistance. Based upon Computers-Are-Social-Actors theory, this study investigates: (1) The way RSRs’ usefulness, social capability, and appearance facilitate Human-Robot Interaction (HRI); (2) whether anxiety toward robots inhibits the relations between RSRs’ facilitators and HRI, and (3) whether HRI affects anticipated service quality and ultimately the acceptance of RSRs. The study incorporates interviews, video clip stimuli, and empirical data collection (n = 1362) in fashion, technology, and food-service. The results reveal that RSRs’ facilitators such as usefulness, social capability, and appearance influence attitudes toward HRI positively, which in turn, predict anticipation of better service quality and greater acceptance of RSRs, and demonstrate the moderating role of anxiety toward robots between RSRs’ facilitators and attitudes toward HRI. The findings provide theoretical and practical implications for the adoption of RSRs and development of effective HRI.

User acceptance mechanism and usage promotion strategy of AI services based on mind perception theory

User acceptance mechanism and usage promotion strategy of AI services based on mind perception theory

디지털 기반 언택트 서비스 도입으로 인한 호텔 브랜드 가치 연구

Based on the expectation confirmation model, this study analyzed the relationships among AI service characteristics, choice confirmation, vicarious satisfaction, self satisfaction, and hotel brand value. In order to test the research model, this study conducted a survey with hotel guests of five-star hotels that implemented AI service in the Republic of Korea. A total of 384 valid surveys were analyzed through structural equation modeling. The results showed that two functional characteristics of the AI service (i.e., convenience and enjoyment) positively affected the choice confirmation of the hotel guests. Results also indicated that a social characteristic of the AI service (i.e., new normal receptivity) positively affected the choice confirmation of the hotel guests. Findings also showed that the psychological characteristics of AI service (i.e., digital efficacy) positively affected the choice confirmation of the hotel guest and, in turn, choice confirmation positively affected both vicarious satisfaction and self satisfaction. Finally, both vicarious satisfaction and self-satisfaction positively affected hotel brand value. Theoretical implications of the study and marketing strategies to management for the AI service of luxury hotel rooms were suggested at the end of the study based on the findings of the study.

2PN: A Unified Panoptic Segmentation Network with Attention Module

Comprehensive and accurate surveillance of the environment forms the basis of secure Internet of things (IoTs), the threats can be observed, and the AI services of IoT systems can be preserved. Panoptic segmentation is an efficient and popular approach for environmental surveillance based on images captured by smart sensing devices. This approach can jointly detect stuffs and things within an image and feed subsequent tasks like image detection. So far, there are many methods for panoptic segmentation which focus on extracting sophisticated visual features for segmentation. However, these efforts are both heavy on their workload and cannot clearly distinguish essential features useful for surveillance in an open environment. Therefore, this paper proposes a novel deep learning model 2PN for panoptic segmentation. The model includes a 2-way pyramid network and an attention module to learn in a more concentrated and reasonable way which enhances the feature extraction part. It strikes a balance between the computing complexity and the power of model capability. Finally, 2PN (2-way pyramid network) results are reflected on the Cityscapes dataset.

A method for reducing cloud service request peaks based on game theory

The rapid development of Internet of Things (IoT) technology and the popularity of Artificial Intelligence (Al) technology research have brought new opportunities for the development of cloud computing (CC). With the increasing number of mobile Internet access devices and IoT access devices, the number of task requests from CC customers for AI services in the network has also experienced an explosive growth. In this paper, the focus is on the possible overload of cloud providers during the peak period of cloud service requests. Thetime attributes of cloud task execution are classified to avoid overloading the cloud provider as much as possible. In a distributed cloud environment, it is necessary to consider the time flexible attributes of cloud tasks to reasonably compete for cloud resources. In this work, game theory (GT) is introduced to formulate a cloud service scheduling game, in which participants are cloud customers who participate in the purchase of cloud services. The players' strategies are the time flexibility of each cloud task. The problem is formulated as minimizing the cost of scheduling cloud services and a noncooperative game among the customers (as players) is presented. Then the existence of the Nash equilibrium (NE) solution of the game has been proved and a new algorithm has been proposed in this paper to compute it. In addition, the analysis process of the convergence of the proposed PCA algorithm and the proof of its convergence to NE are also included in this paper. At the end of the paper, simulations were performed to verify the theoretical analysis presented. The experimental results show that the proposed PCA algorithm can converge to the Nash equilibrium very quickly, effectively reducing the peak value and increasing profit.

Information Retrieval for Aviation Applications

<div class="section abstract"><div class="htmlview paragraph">There is a recurring need for automatic Information Retrieval (IR) from quality documents, price tags, part markings, receipts, purchase orders and technical manuals - which are otherwise non-parsable. IR coupled with search functionalities has a wide range of applications from warehouses, marketplaces to even shop floors in the aviation sector. It helps in semi-automating workflows like document reviews, quality checks, collaborative Q&As and contextual extraction of information. These workflows make laborious tasks more intuitive and easier, thereby reducing the workload of the engineers using them. The paper describes an AI based IR platform which caters to the aforesaid scenarios in a scalable manner and integrates seamlessly with similar problems across different domains. It is the core to many different workflows that are currently used to detect paperwork mismatch for aviation parts, to auto-catalogue large amounts of documents, to detect the presence of sensitive information in documents, to digitize scanned documents, detecting presence of stamps in faceplates. Its capabilities can be extended beyond documents to other non-interactive and non-parsable contents as well. The intent of this paper is to describe different AI modules and services in this AI platform towards enabling Smart Factory and Industry 4.0.</div></div>