Data Science

Abstract
This paper introduces a novel algorithm designed to enhance the performance of PDF-based chatbots for domainspecific
question answering. The proposed system integrates advanced table parsing techniques, hybrid indexing,
context-aware response generation, and a continuous learning feedback loop, effectively addressing the limitations
of existing approaches. We validate the effectiveness of our algorithm through a case study using a PDF document
on pregnancy, demonstrating its potential applications across various specialized knowledge domains. The
chatbot’s innovative features, such as advanced multimodal content processing and dynamic knowledge base
updates, establish it as a powerful tool for extracting relevant information from complex documents.

The prescriptive agile framework is a novel approach that leverages an integrated process ontology, involving agile Scrum and OpenUp methodologies. The objective is to enhance both productivity and quality within software development projects where fresh developers are engaged. Key Process elements of OpenUP and agile Scrum are used to construct the ontology. The ontology is defined and instantiated with software projects having the complete user stories. Subsequently a three-sprint project experiment is conducted. The metrics of Earned Value (EV) and Code Maintainability are used to gauge productivity and quality. This research contributes to improving productivity and code quality in software engineering particularly when there are many amateurs.

Performance of individual developers and effort on reworks in agile scrum environments has an impact on the productivity and quality of the developed software. The traditional agile metrics do not consider reworks in measuring productivity. Yet, an enhanced methodology to compute individual performance and quantify reworks in agile scrum contexts is required. By introducing Earned Value Analysis based Productivity Ratio and Blended Quality Metric based on Code Maintainability Index, this gap is bridged. Applying the metrics to ten real time software projects over eighteen months, having twelve sprints each, a framework is decided for performance management in agile scrum development. The findings indicate the efficacy of the metrics in providing valuable insights to improve productivity and quality measures in software development environments.

Internet of Things (IoT) applications have become ubiquitous in various aspects of daily life, including smart homes, healthcare, and other areas where human assistance is crucial. However, the valuable real-world data collected by IoT devices and transmitted over the Internet have become a prime target for numerous malicious actors and hackers. Therefore, ensuring secure communication to prevent unauthorized access to this transmitted data is of paramount importance. This research is dedicated to the development of a robust security system for IoT to protect sensor data. Traditionally, IoT systems have relied on fixed encryption keys shared between transmitters and receivers, which presented vulnerabilities as these keys could be easily compromised. In this study, we introduce an enhanced version of the AES algorithm with dynamic keys. In this modified algorithm, each cycle involves the XOR operation of four keywords with the sequence number. Each round of the AES algorithm incorporates various transformations, including DivideSwap, SubBytes, ShiftRows, MixColumns, and AddRoundKey, with the exception of the final round, which excludes the MixColumns operation. To implement and test this upgraded security approach, we utilized hardware components such as the Arduino Uno, ESP8266, and DHT11 sensors. The enhanced AES algorithm was integrated into the Arduino Uno to secure sensor data before transmission across the network. The results of our analysis demonstrate that the proposed IoT security method presents a significantly higher level of cryptographic resilience compared to the traditional AES algorithm, making it a robust solution for safeguarding IoT data.

The Internet of Things is a developing network of embedded gadgets that can connect with the outside world (IoT). Since IoT devices have limited storage, memory, and processing power, they should typically offer the choice to run at lower power levels. These constrained devices are unsuited for security measures that significantly rely on encryption since they cannot carry out complex encryption and decryption quickly enough to allow for the possibility of reliably transferring information in real-time. A public-key encryption method called elliptic curve cryptography is utilised in an IoT setting to guarantee security, however it takes a long time to evaluate because of its scalar increase. While still adhering to the requirements of extremely lightweight encryption, ECC is currently slower than AES. The memory requirements for ECC must be decreased in order for the system to consume less RAM and ROM during operation. This work decreases the scalar increase in the elliptic curve and proposes a novel mapping strategy in order to speed up the assessment process. The succession number multiplies the plain text corresponding ASCII values. The elliptic curve will be input with the multiplied values. By using this innovative technique, ASCII value plaintext avoids overt repetition. The technique is intended to be used to encode or decode any plaintext message that contains ASCII values and a sequence number.

Elliptic Curve Cryptography (ECC) is a current area of enquiry in the domain of Cryptography. ECC is beneficial in various schemes of cryptography due to its lowered key size and rapid generation of key. It is suitable for IoT devices to provide security. Many IoT devices have restricted amount of storage and processing ability. They frequently need to run on low power. Heavy encryption dependent approaches are not a good fit for these constrained devices. They are incapable of performing complicated encryption and decryption quick enough to be able to secure data in real-time transmission. Transformation of text messages to elliptic curve points has always been enigmatic. Comparing to other cryptographic algorithms the arithmetic involved in elliptic curve cryptography is computationally less complicated. This paper provides an overview of elliptic curves, different mapping schemes and their application in cryptography. The emphasis is on the performance gains to be acquired by using elliptic curve cryptography as an alternative to conventional cryptosystem like RSA.

The implicit idea of the IoT is to follow client's identity effectively, so the security and protection concerns like theft the information and disturbance of tasks are turning out to be basic issues in the present IoT applications. Aggressor can pick up the access to the system can harm the physical gadgets of IoT can harm the system which bargains the protection and security of IoT. As the IoT gadgets have low memory, less force and data transmission a productive arrangement to give security is required that won't chomp through the resources of IoT. Because of the resource constrained condition in IoT the regular algorithms isn't sufficient to guarantee the information security. So we want a less computational expense as far as force utilization, memory management and more effective cryptography algorithms to offer more information security in IoT. This paper shows the hybrid cryptography security in IoT by embedding of ECC and TwoFish calculation, which gives an enhanced answer for upgrade the security highlights of the IoT with the goal that we can improve the administration therefore achieves growing the trust over the advancement.

The Internet of Things (IoT) is a paradigm where everyday objects can be equipped with
identifying, sensing, networking and processing capabilities to communicate with one another over
the Internet. There are various healthcare monitoring systems available now which uses internet of
things components. This paper surveys on IoT-based health care monitoring systems available and
their working methodologies. It is observed from the survey that most of the IoT healthcare systems
employ various sensors for reading data from the patients, and then the data transfer to cloud through
Base station/Access points, and finally the data analysis and interpretation by specialists. This paper
reviews various recent improvements in the field and outlines some of the challenges in IOT based
healthcare monitoring systems.

Elliptic Curve Cryptography is a novel exploration public key algorithm in the region of Cryptography. It is reasonable for IoT gadgets to give security. Numerous IoT gadgets have confined measure of processing ability and memory capacity. Elliptic curve furnishes more significant level of security with less key size contrasted with other lightweight Cryptographic methods. Change of instant messages to elliptic curve points has been baffling all of the time. A new mapping scheme has been proposed in this paper. The relating ASCII values of the plain text are multiplied with sequence number. The multiplied values will be input for the elliptic curve. This new method evades the overt repetitiveness ASCII value plaintext. The algorithm is designed so that it tends to be utilized to encode or decode any given plaintext message of ASCII values with sequence number.

ABSTRACT
Agriculture is first and foremost factor which is important for survival. Machine learning (ML) could
be a crucial perspective for acquiring real-world and operative solution for crop yield issue.
Considering the present system including manual counting, climate smart pest management and
satellite imagery, the result obtained arent really accurate. This paper focuses mainly on predicting
the yield of the crop by applying various machine learning techniques. The classifier models used
here include Logistic Regression, Naà ̄ve Bayes and Random Forest, out of which the Random
Forest provides maximum accuracy. The prediction made by machine learning algorithms will help
the farmers to come to a decision which crop to grow to induce the most yield by considering factors
like temperature, rainfall, area, etc. This bridges the gap between technology and agriculture sector.
Predicting crop yield based on the environmental, soil, water and crop parameters has been a
potential research topic. Deep-learning-based models are broadly used to extract significant crop
features for prediction. Though these methods could resolve the yield prediction problem there exist
the following inadequacies: Unable to create a direct non- linear or linear mapping between the raw
data and crop yield values; and the performance of those models highly relies on the quality of the
extracted features.
KEYWORDS Crop_yield_prediction, Logistic_regression, Naà ̄ve bayes,Random forest,
Weather_api, Climate change, deep learning, Precision algorithm

The process of extracting information to identify patterns, trends, and useful data
that would allow the business to take the data-driven decision from huge sets of data is called
Data Mining. Data Mining is the process of investigating hidden patterns of information to
various perspectives for categorization into useful data, which is collected and assembled in
particular areas such as data warehouses, efficient analysis, data mining algorithm, helping
decision making and other data requirement to eventually cost-cutting and generating revenue.
Data mining is the act of automatically searching for large stores of information to find trends
and patterns that go beyond simple analysis procedures. Data mining utilizes complex
mathematical algorithms for data segments and evaluates the probability of future events. Data
Mining is also called Knowledge Discovery of Data (KDD).This process includes various types
of services such as text mining, web mining, audio and video mining, pictorial data mining, and
social media mining. The biggest challenge is to analyze the data to extract important
information that can be used to solve a problem or for company development. There are many
powerful instruments and techniques available to mine data and find better insight from it.
Keywords
ï‚· Data mining
ï‚· Data warehouse
ï‚· Pattern
ï‚· Data cleaning
ï‚· Knowledge extraction
ï‚· Transformation
ï‚· Evaluation