65 research outputs found
Human capital, fertility and growth under borrowing constraints
No full textIn this paper we investigate economic growth in economies where households face liquidity constraints, and young agents rely on the family to finance their investments in education. We analyze the type of family aid in which youths can borrow because their parents guarantee the loan repayment with their income. In an OLG model of economic growth, it is shown how multiple equilibria can arise. A stable trap of low-development is characterized by high fertility rates and low investment in human capital. On the other hand, economies with a sufficiently low starting rate of fertility grow according to a process that may describe a demographic transition. In this case, borrowing constraints gradually vanish and the process of growth reaches a steady state characterized by the optimality of fertility and schooling choices. Econometric evidence on the significant roles of family income and size, and credit constraints among the determinants of international secondary school enrollment rates is provided to support the main hypotheses of the model.OLG, Human capital, Multiple equilibria
Hate Speech Detection Using Textual and User Features
No full textSocial media platforms provide users with a powerful platform to share their ideas. Using one’s right to expression to incite hatred toward a particular group of people is inappropriate. However, hate speech is pervasive in our society. Spreading hate through online social networks like Facebook, Twitter, Tiktok, and Instagram is commonplace in today’s milieu. One such case is the unprecedented COVID-19 pandemic, which engendered anti-Asian hate.
In current literature, there is limited study on using user features in conjunction with textual features to detect hate. This thesis aims to combine textual features with user features to improve the state-of-the-art hate speech detection technique. To test our approach, we used four different datasets available in the public domain. We have used various tools to access Twitter APIs to extract required user information, either to use directly or further compute other features using that information.
We have represented the textual features in the form of BERT embeddings and linguistic features. The 97 linguistic measures computed with a Linguistic Inquiry and Word Count (LIWC) tool quantify the text’s cognitive, affective, and grammatical processes. The user feature consisted of demographic, behavioral-based, emotion-based, personality, readability, and writing style features. Our experimental evaluation over three datasets shows that the top twenty linguistic features and the top twenty user features are the best combinations for hate speech detection.
Hate speech is mostly emotionally charged. We further analyzed these user and linguistic features. Among the most intuitive and prominent results was that features like anger, negative emotion, swearing, fear, and annoyance were high in hate speech, while the happiness feature was low.
We compared multiple approaches along with the existing state-of-the-art. We found that the best approach with textual features was combining LIWC features with BERT embeddings. This combination gave us the F1 of 0.82 and 0.79 on Crowd-sourced (DS1) and Kaggle (DS3), respectively. Followed by this, we identified the top LIWC and user features for hate speech detection. We found that features representing negative emotions like anger, fear, sadness, and annoyance were prominently high in hate speech. Happiness is lower in hate speech. After this, we analyzed the F1 scores with standalone LIWC and user features. We also used their combinations. We found that the combination of the top twenty LIWC and top twenty user features gives the best F1 scores of 0.74, 0.90, and 0.64 on DS1, NAACL (DS2), and anti-Asian Covid hate (DS4) dataset.
Finally, we used traditional machine learning algorithms combining BERT embeddings with the top twenty linguistic features and the top twenty user features. We obtained the F1 scores of 0.78, 0.92, and 0.84 on DS1, DS2, and DS4 respectively. We also compared our approach with other studies using user and textual features
Raut recognized with award for outstanding contributions to service-learning through research
No full text
Semi-custom VLSI Design and Realization of DC-DC Converters in UMC90
No full textAs CMOS technology is scaling down, the effect of voltage drop in power distribution network is becoming more prominent. Such voltage drops on power lines in a clock network introduces significant amount of skew, thereby degrading the signal integrity. With rising power consumption and decreasing supply voltages, the upply current will increase in future devices. The IR drops can be reduced by using large wires which negatively impacts the global routing. Thus to provide proper supply voltages, on chip DC-DC converters are designed. The purpose of this project is to design an on-chip DC-DC converter targeted for System on Chip (SOC). In this thesis switched capacitor up converter and differential based voltage down converter is designed in UMC90. A new design for differential based voltage down converter is described to increase the efficiency of the converter. The specifications of the converters are defined by the system requirements. After meting the system requirements, layout of both converters is designed. The converters designed have high efficiency and small layout area.Microelectronics & Computer EngineeringElectrical Engineering, Mathematics and Computer Scienc
A Review of Question Answering Systems: Approaches, Challenges, and Applications
No full textQuestion answering (QA) systems are a type of natural language processing (NLP) technology that provide precise and concise answers to questions posed in natural language. These systems have the potential to revolutionize the way we access information and can be applied in a wide range of fields including education, customer service, and health care.There are several approaches to building QA systems, including rule-based, information retrieval, and machine learning-based approaches. Rule-based systems rely on predefined rules and patterns to extract answers from a given text, while information retrieval systems use search algorithms to retrieve relevant information from a large database. Machine learning-based systems, on the other hand, use training data to learn to extract answers from text.One of the main challenges faced by QA systems is the need to understand the context and intent behind a question. This requires the system to have a deep understanding of the language and the ability to make inferences based on the given information. Another challenge is the need to extract relevant information from a large and potentially unstructured dataset.Despite these challenges, QA systems have a wide range of applications, including education, customer service, and health care. In education, QA systems can be used to provide personalized learning experiences and help students learn more efficiently. In customer service, QA systems can be used to handle a high volume of queries and provide quick and accurate responses to customers. In health care, QA systems can be used to assist doctors and patients by providing timely and accurate information about medical conditions and treatments.Overall, this review aims to provide a comprehensive overview of QA systems, their approaches, challenges, and applications. By understanding the current state of development and the potential impact of QA systems, we can better utilize these technologies to improve various industries and enhance the way we access information
Personalize News Recommendation System by Using Stack Auto Encoder
No full textThe popularity of Internet and mobile Internet, people are facing serious information overloading problems now a days. Recommendation engine is very useful to help people to reach the Internet news they want through the network. Recommender Systems have become the vital role in recent years and are utilized widely in various areas of social importance. In day to day life, users will not be able to read news every day due to heavy schedule. So to increase General knowledge of users we propose online recommendation systems which recommend news. In existing system i.e. news delivery portals deliver popular news on home page of the portal but user’s data according recommendation is not implemented yet. To overcome existing system problems we propose new recommendation system which automatically finds the news based on user’s profiles. Using the stack auto encoder algorithm
- …
