Description:Field of the Invention
The proposed invention is to revolutionise student assistance with Artificial Intelligence with special emphasis on natural language processing and conversational systems, by intaking the student's queries regarding the CSE-AI&ML department and ensuring to provide better suggestions with the help of a well-tuned Large Language Model(LLM).
Objectives of the Invention
The main objective of the revolutioning idea EduAI: Chatbot for Student Assistance, is to provide an AI-driven solution supported by Natural Language techniques and assists students in departmental affairs. This invention aims to intake the user-defined queries from the students, modify the queries to process and provide better results to assist them. The invention provides access at any time and is performed through automation, this provides better accurate outputs and reduces the manual efforts for resolving their queries. The model enhances the structure, tone and modulation of the text effectively, such that it can provide a well-developed and prompting output with it.
Background of the Invention
AI-Powered Chatbots for Student Departments in Colleges: Few decades ago, with usage the traditional human-to-human interaction for academic advising; a large population of students were not able to interact to the respective faculty for the resolution of academic queries along with the extensive wait times, resources and inefficient navigation through vast academic catalogues. Due to above issues, it had been affecting students with several challenges like graduation plans. To avoid all these problems, AI-based chatbots provide a better context-specific outputs with the help of Natural Language techniques. Such systems are being connected to the institutional databases to provide better outputs and be as precise as possible, by being most recent data to be able to extracted and used to answer the queries. Over time by training, the system supports adaptability and becomes better with contextual understanding and queries being asked. Due to this, AI-based chatbots can contribute significantly to the effectiveness of academic advising through real-time support, a drastic decrease in workload over the human advisors, and in turn, student can pose queries with the support available. The system is also a scalable solution for the growing needs of the modern educational institutions and making them leverage AI for this aspects.
US20220129799A1 is a real-time guidance system that provides the students with the academic policies, credit transfer processes and graduation requirements. This system connects with institutional databases to provide them, with the context-specific interactions being tailored to provide better results. The System enables the student who can interact and have a clear-understanding over the process involving in selection of courses, fulfilling credits, and the requirements for graduation based over the data being presented on courses, credit hours and institutional requirements. It would ensure provide timely and proper advisement so experience is better with the system since much academic information becomes available without compromising the support. It grants student in deciding what serve them best in pursuit. It provides, in real-time dependence on human advisors and increases overall general rates of academic success of the students.
US20230167345A1, is a system that provides dynamic academic support that is being powered by an AI chatbot for use by the students of a college to understand the complexities of the college requirements, deadlines and available academic resources. This system leverages machine learning and natural language processing to increase its ability to answer those questions that are being asked by the students based on experience. The system continuously learns from each of its conversation with the students and to attain better results, it should be treated more with the training that gives the ability to provide with good results. The chatbot helps students by providing personlized and real-time support to those who need help in their academic enquiries in choosing course selection or graduation requirements. In this respect, the system enhances meeting the needs of students by equipping the student with proper and timely information and timely improves with its academic experience.The current invention US20210039126A1, discusses about the student support system which provides students with all the information related to courses and schedules, deadlines; by provideing students with all the necessary information and materials required for the academics. The AI-based chatbot uses the machine learning models to process queries from students before answering them in an appropriate manner. Since, many of the queries of the students would be similar and the chatbot would be able provide with the information based on its experiencce being in a school. This will reduce the manual workload and as most of questions being answered for the repetitive question by computers automatically. Therefore, it streamlines procedures so that can have the information they require in time. As the interaction with the students to the chatbot increases, chatbot learns and evolves in terms of producing more accurate responses over time.US20230174675A1 is an innovative chatbot system that is being supported by AI, used for tailored responses for higher education institutions. The primary aim of the system to assist students with academic needs such as course offerings, enrollment process. The system uses the input query as the inputs and analyze it with the help of natural language processing. Using the machine learning models, the system tries to implement the answers with better manner with the help of adaptive learning. This takes care of the student's academic queries till his/her graduation. This approach streamlines the advice procedure and also provides the better answers regarding institutional inquiries which are more complex to deal. This invention provides, context-specific, tailored assistance to the students, and it highlights a scalable solution for improving the efficiency of academic advising while enhancing the overall student experience of UI as well.
US20230156719A1 is an innovative system that deals with the assistance of course selection and academic planning using AI. The primary goal of this system is to be integrated into institutional databases and provide personalised and tailored advice. It will understand the prerequisites, course completion details and elective selection process, guiding them in making the best decisions possible to support their academic and career goals. The invention extends to usage for being as the assistant that provides specific monitoring of individual student progress over time to employ analytics, in determining the possible gaps and encourage the students to be on path for graduation. The invention leverages the machine learning techniques that provide the best possibilities for adaptive learning. This dynamic form of adaptability guarantees that the system evolves along with the changing institutional policies and student queries as well.
Summary of the Invention
EduAI is a revolutionary AI chatbot that facilitates college student's access to academic information. It works like an intelligent assistant answering questions related to the curriculum, course information, and requirement-specific semester details in quick and accurate ways. By scanning huge number of documents and summarizing the information into answers, the AI relieves students of the tedious manual search that students can access direct and specific answers. Therefore, EduAI is a helpful navigator on the guidance path for academic life.
At the heart of EduAI's functionalities, the strongest equipment is found in the Llama 3.1 large language model. Using the 'all-MiniLM-L6-v2' embedding model, it is also responsible for really fast-indexing and recovering of relevant information from large data sets. In fact, this combination of features leads to speed generation of responses, making such a chatbot reliable for a variety of academic queries to answer. The design of EduAI also maximizes resources and allows students to invest more of their time in studies, lack of consumption in finding information.
EduAI is a major development in AI integrated into education. It should make one's life easier by having to put in minimal effort while helping tailor service specifically to the different needs of various users. It could just be a matter of understanding the course syllabus and semester's requirements, but EduAI handles it all. The streamlined approach and advanced technology regarding the EduAI have the power to make it an inevitable resource for solving all academics-related problems for students.
Brief Description of Drawings
In the following, we shall describe the invention in great detail with respect to the particular embodiments shown in the figures in particular:
Figure-1: Detailed architecture of the proposed invention.
Figure-2: Basic Workflow diagram of the proposed invention.
Detailed Description of the Invention
The following invention "EduAI: Chatbot for Student Assistance" generates an AI-driven chatbot as a solution to support academic guidance for students. This system uses the machine learning models like "all-miniLM-v2" for indexing of the files that are being provided as inputs; while Facebook AI Similarity Search(FAISS) is being used for more effective and efficient chunk retrieval. It uses Meta's recently launched Llama 3.2, with its generative capabilities to generate context-specific responses to queries being asked by the students. The user interface is being developed with Streamlit, providing an easy-to-operate and user-friendly interface.The "all-miniLM-v2" model is a lightweight and efficient transformer-based language model and plays an prominent role in embedding the institutional-related data into higher dimensional vectors. These embeddings provide effective and efficient indexing and search operations using FAISS. FAISS is being used for quick similarity checks and making it possible to extract the information that has contextually relevant information with less delay in time. This adds to the capability of implementing high-speed query processing with large datasets, like course catalogs, schedules and academic policies.FAISS indexing is being integrated to manage vast amounts of student-related data. The system utilizes FAISS in such a way that indexes can be searchable, allowing the chatbot to query efficiently over the datasets. FAISS provides the chatbot retrieves the accurate and tailored answerd from the Large Language Model when the user seeks any information about course prerequisties. This indexing enables the scalability of the application as new academic inputs are fed to it.
Llama 3.2 is a state-of-the-art language with natural language understanding and contextual output generation, used for the actual output generation. This makes replies from EduAI precise, relevant and contextual in answering every query being faced by students. The capability of answering nuanced questions, such as related course credits and very specific departmental policies, makes this model, a perfect choice for academic assistance.The base framework upon which the interface of EduAI is built is Streamlit. It is supposed to be simple and easy to use;it is aimed to provide ease to use for different stakeholders including students. Streaming from it is so easy and informl, allowing the users to easily chat with the chatbot and get immediate qnswer to queries. It promises much with its clear, neat and design; to engage a conceivable touch of students interests while providing great user experience to the student in need of academic assistance.Combining the 'all-miniLM-v2' model, FAISS and Llama 3.2 parameters enhances a balance between systems-fast, accurate and scalable. Combining FAISS's indexing with Llama's contextual response generation paves the way for a positive effect on students through a process of custom, dynamic assistance by the chatbot. This symbiosis would conform to the dynamics of changing datasets and will respond to the needs of students over a significant time.EduAI deals with the flexibility and integrate to numerous academic ecosystems and databases. The model intakes the institution-specific policies, program details and academic calendars into its database. The mobility in the chatbot remains relevant and helpful for the students across various educational institutions. EduAI provides a better scalability for the underlying technologies to handle increasing data volumes without any reduction in the performance.
The invention provides security to data and privacy by making personalised chats with each user separately and maintaining the student information in private.FAISS indexing is generally combined with encryption techniques to avoid unknown access to the chatbot. Besides this, the Llama 3.2 model performs its operation within the boundaries of security only with privacy checks. In this way, the security increases trust in the chatbot resulting in EduAI as a reliable tool for academic assistance. The chatbot supports a notable feature that has the ability to learn with the prior interactions with the students and adapting the kind of feature to be added to the chatbot through learning with those interactions. The upgrades in learning process makes the system more efficient in various aspects and also provides with improved assistance with the new queries of students. Also, being able to attain experience over newly updated data as well. With the following features keeps EduAI, as a better solution for academic assistance.The proper setup completion along with the deployment of model of EduAI, adds to the effective usage of the automation in academic assistance. It minimizes the dependency over human advisors in academic assistance and reduces the administrative workloads as well. Along with it, it provides better support to studensts. With all these features being supporting the invention and providing better efficiency for the academic institutions to focus over improving educational outcomes and the students should also have the access to reliable assistance. The invention 'EduAI: Chatbot for Student Assistance', plays a key role in connecting students and educational institutions. Such an invention makes use of a combination of models such as 'all-miniLM-v2' to generate the FAISS index and provides extensive support of Llama 3.2 too for making a highly efficient and user-friendly solution for academic assistance. With integration of Llama 3.2 along with Streamlit ensures easy accessing and best possible architecture that results in better scalability and adaptability. EduAI supports with the diverse needs of students in assistance and provides them with a proper explanation and it makes a prominent advancement in the domain of educational technology and AI.
3 Claims and 2 Figures

Equivalents
The present invention, a Large Language model for Student Assistance based on Natural Language Processing tools, provides better outcomes with higher efficiency. The model intakes the queries of the user and is used for standardizing it to identify the better result for the query being given. Because of its adaptability, it can be able to reduce manual efforts along with time consumption. , Claims:The scope of the innovation is defined by the following claims:

Claim:
1. The invention EduAI: Chatbot for Student Assistance comprising,
a) The system integrates with Natural Language techniques to analyze the query provides a better refined version of it and generates the output for the query.
b) The chatbot employs the adaptive learning technique for understanding frequently asked queries and supports better responsiveness.
c) The invention uses the semantic similarity algorithm to match user queries, retrieve the answers and tune before the generation of answers.
2. According to claim 1, the Natural Language Processing tools, such as the removal of stop words, etc., provide better analysis of the query done and also support the elimination of stop words and perform better search over the vector database.
3. As per claim 1, the semantic similarity algorithm is integrated to measure the relevance of user queries over the vector database. By retrieving the most similar data, the algorithms also tune the information that is being extracted from the database to provide better results.