app.py

import streamlit as st
import torch
from transformers import AutoTokenizer, AutoModelForSequenceClassification
import fitz
import os

model = AutoModelForSequenceClassification.from_pretrained("Reem333/Longformer")
tokenizer = AutoTokenizer.from_pretrained("allenai/longformer-base-4096")

def extract_text_from_pdf(file_path):
    text = ''
    with fitz.open(file_path) as pdf_document:
        for page_number in range(pdf_document.page_count):
            page = pdf_document.load_page(page_number)
            text += page.get_text()
    return text

def predict_class(text):
    try:
        max_length = 4096
        truncated_text = text[:max_length]

        inputs = tokenizer(truncated_text, return_tensors="pt", padding=True, truncation=True, max_length=max_length)
        with torch.no_grad():
            outputs = model(**inputs)
            logits = outputs.logits
            predicted_class = torch.argmax(logits, dim=1).item()
        return predicted_class
    except Exception as e:
        st.error(f"Error during prediction: {e}")
        return None

uploaded_files_dir = "uploaded_files"
os.makedirs(uploaded_files_dir, exist_ok=True)

class_colors = {
    0: "#d62728",  # Level 1
    1: "#ff7f0e",  # Level 2
    2: "#2ca02c",  # Level 3
    3: "#1f77b4"   # Level 4
}

st.set_page_config(page_title="Paper Citation Classifier", page_icon="logo.png")

with st.sidebar:
    st.image("logo.png", width=70)
    st.markdown('<div style="position: absolute; left: 5px;"></div>', unsafe_allow_html=True)
    
    st.markdown("# Paper Citation Classifier")
    st.markdown("---")
    st.markdown("## About")
    st.markdown('''
    This is a tool to classify paper citations into different levels based on their number of citations.
    Powered by Fine-Tuned [Longformer model](https://huggingface.co/Reem333/Longformer) with custom data.
    ''')
    st.markdown("### Class Levels:")
    st.markdown("- Level 1: Low cited papers")
    st.markdown("- Level 2: Average cited papers")
    st.markdown("- Level 3: More cited papers")
    st.markdown("- Level 4: Highly cited papers")
    st.markdown("---")
    st.markdown('Tabuk University')

st.title("Check Your Paper Now!")

#option = st.radio("Select input type:", ("Text", "PDF"))

#if option == "Text":
title_input = st.text_area("Enter Title:")
abstract_input = st.text_area("Enter Abstract:")
    
affiliations_input = st.text_area("Enter Affiliations:")
keywords_input = st.text_area("Enter Keywords:")
options=['Environmental Sciences; Remote Sensing', 'Construction & Building Technology; Energy & Fuels; Engineering, Civil', 'Energy & Fuels', 'Chemistry, Physical; Materials Science, Multidisciplinary; Metallurgy & Metallurgical Engineering', 'Multidisciplinary Sciences', 'Nursing', 'Chemistry, Applied; Food Science & Technology; Nutrition & Dietetics', 'Medicine, General & Internal', 'Computer Science, Artificial Intelligence; Computer Science, Software Engineering; Computer Science, Theory & Methods; Engineering, Electrical & Electronic; Optics', 'Pharmacology & Pharmacy; Toxicology', 'Surgery', 'Biodiversity Conservation; Environmental Sciences', 'Engineering, Chemical', 'Infectious Diseases; Microbiology', 'Neurosciences', 'Environmental Sciences; Meteorology & Atmospheric Sciences', 'Engineering, Environmental; Environmental Sciences', 'Health Care Sciences & Services; Public, Environmental & Occupational Health', 'Pharmacology & Pharmacy', 'Agriculture, Multidisciplinary', 'Environmental Sciences', 'Cardiac & Cardiovascular Systems', 'Endocrinology & Metabolism', 'Environmental Studies', 'Forestry', 'Behavioral Sciences; Nutrition & Dietetics', 'Computer Science, Information Systems; Health Care Sciences & Services; Medical Informatics', 'Computer Science, Software Engineering; Computer Science, Theory & Methods', 'Geochemistry & Geophysics', 'Parasitology; Tropical Medicine', 'Fisheries', 'Thermodynamics', 'Environmental Sciences; Public, Environmental & Occupational Health', 'Business', 'Plant Sciences; Chemistry, Medicinal; Integrative & Complementary Medicine; Pharmacology & Pharmacy', 'Medicine, Research & Experimental; Pharmacology & Pharmacy', 'Engineering, Multidisciplinary; Mathematics, Interdisciplinary Applications; Mechanics', 'Materials Science, Coatings & Films; Physics, Applied', 'Oncology; Radiology, Nuclear Medicine & Medical Imaging', 'Food Science & Technology', 'Biochemistry & Molecular Biology; Biotechnology & Applied Microbiology', 'Engineering, Mechanical; Materials Science, Characterization & Testing', 'Biology; Mathematical & Computational Biology', 'Chemistry, Multidisciplinary', 'Public, Environmental & Occupational Health; Infectious Diseases', 'Nutrition & Dietetics', 'Mathematics, Applied; Mathematics, Interdisciplinary Applications; Mechanics; Physics, Fluids & Plasmas; Physics, Mathematical', 'Medicine, Legal; Social Sciences, Biomedical', 'Biochemistry & Molecular Biology', 'Infectious Diseases', 'Spectroscopy', 'Agriculture, Dairy & Animal Science', 'Business, Finance', 'Psychology, Developmental; Neurosciences', 'Public, Environmental & Occupational Health', 'Thermodynamics; Engineering, Mechanical; Mechanics', 'Ecology; Environmental Sciences', 'Geochemistry & Geophysics; Meteorology & Atmospheric Sciences', 'Oncology; Urology & Nephrology', 'Geography, Physical; Geosciences, Multidisciplinary; Remote Sensing; Imaging Science & Photographic Technology', 'Psychiatry', 'Chemistry, Physical; Materials Science, Coatings & Films; Physics, Applied; Physics, Condensed Matter', 'Computer Science, Interdisciplinary Applications; Education & Educational Research', 'Environmental Studies; International Relations', 'Nuclear Science & Technology', 'Computer Science, Software Engineering', 'Radiology, Nuclear Medicine & Medical Imaging', 'Infectious Diseases; Pharmacology & Pharmacy', 'Surgery; Transplantation', 'Business, Finance; Economics', 'Clinical Neurology', 'Development Studies; Environmental Studies; Regional & Urban Planning; Urban Studies', 'Materials Science, Multidisciplinary; Metallurgy & Metallurgical Engineering', 'Automation & Control Systems; Computer Science, Artificial Intelligence; Robotics', 'Urology & Nephrology', 'Biochemistry & Molecular Biology; Immunology', 'Green & Sustainable Science & Technology; Engineering, Environmental; Environmental Sciences', 'Statistics & Probability', 'Cell & Tissue Engineering; Biotechnology & Applied Microbiology; Cell Biology', 'Materials Science, Multidisciplinary; Physics, Multidisciplinary', 'Food Science & Technology; Nutrition & Dietetics', 'Engineering, Industrial; Ergonomics; Psychology, Applied', 'Ophthalmology', 'Marine & Freshwater Biology; Toxicology', 'Neurosciences; Peripheral Vascular Disease', 'Energy & Fuels; Engineering, Chemical', 'Materials Science, Characterization & Testing; Polymer Science', 'Geology; Mineralogy; Mining & Mineral Processing', 'Behavioral Sciences; Neurosciences', 'Computer Science, Artificial Intelligence; Neurosciences', 'Obstetrics & Gynecology', 'Acoustics; Chemistry, Multidisciplinary', 'Instruments & Instrumentation; Nuclear Science & Technology; Physics, Nuclear; Physics, Particles & Fields', 'Environmental Sciences; Toxicology', 'Engineering, Electrical & Electronic; Instruments & Instrumentation', 'Architecture', 'Soil Science', 'Economics', 'Metallurgy & Metallurgical Engineering', 'Orthopedics', 'Biochemistry & Molecular Biology; Endocrinology & Metabolism; Pharmacology & Pharmacy', 'Biochemical Research Methods; Biochemistry & Molecular Biology; Chemistry, Analytical', 'Hematology; Peripheral Vascular Disease', 'Mechanics; Materials Science, Composites', 'Computer Science, Artificial Intelligence; Engineering, Electrical & Electronic; Operations Research & Management Science', 'Thermodynamics; Energy & Fuels', 'Agriculture, Multidisciplinary; Food Science & Technology', 'Dentistry, Oral Surgery & Medicine', 'Energy & Fuels; Geosciences, Multidisciplinary', 'Meteorology & Atmospheric Sciences', 'Neurosciences; Neuroimaging; Radiology, Nuclear Medicine & Medical Imaging', 'Surgery; Peripheral Vascular Disease', 'Economics; Regional & Urban Planning', 'Agricultural Engineering; Agronomy', 'Geography; Regional & Urban Planning', 'Allergy; Immunology', 'Biochemistry & Molecular Biology; Biophysics; Cell Biology', 'Engineering, Biomedical; Materials Science, Biomaterials', 'Chemistry, Physical; Physics, Atomic, Molecular & Chemical', 'Nanoscience & Nanotechnology; Physics, Condensed Matter', 'Engineering, Chemical; Engineering, Mechanical; Environmental Sciences; Meteorology & Atmospheric Sciences', 'Acoustics; Radiology, Nuclear Medicine & Medical Imaging', 'Health Care Sciences & Services; Medicine, General & Internal; Clinical Neurology', 'Materials Science, Multidisciplinary', 'Construction & Building Technology; Engineering, Civil', 'Chemistry, Physical', 'Immunology; Medicine, Research & Experimental', 'Chemistry, Applied; Chemistry, Physical; Engineering, Chemical', 'Mathematics, Applied', 'Cardiac & Cardiovascular Systems; Peripheral Vascular Disease', 'Environmental Sciences; Marine & Freshwater Biology', 'Physics, Multidisciplinary', 'Biology; Physiology', 'Materials Science, Biomaterials', 'Environmental Sciences; Environmental Studies; Meteorology & Atmospheric Sciences', 'Green & Sustainable Science & Technology; Energy & Fuels', 'Chemistry, Analytical; Electrochemistry; Instruments & Instrumentation', 'Agricultural Engineering; Biotechnology & Applied Microbiology; Energy & Fuels', 'Hospitality, Leisure, Sport & Tourism; Sociology', 'Computer Science, Artificial Intelligence', 'Agronomy; Plant Sciences', 'Demography; Economics; Gerontology', 'Business; Economics; Management', 'Computer Science, Theory & Methods', 'Water Resources', 'Substance Abuse; Psychiatry', 'Business; Regional & Urban Planning', 'Optics', 'Psychology, Educational', 'Economics; Health Care Sciences & Services; Health Policy & Services', 'Psychology, Experimental', 'Computer Science, Interdisciplinary Applications; Information Science & Library Science', 'Mechanics', 'Environmental Sciences; Environmental Studies', 'Chemistry, Analytical', 'Gastroenterology & Hepatology', 'Geography, Physical; Geosciences, Multidisciplinary', 'Clinical Neurology; Pediatrics', 'Immunology; Neurosciences; Psychiatry', 'Parasitology', 'Engineering, Chemical; Polymer Science', 'Infectious Diseases; Microbiology; Respiratory System', 'Construction & Building Technology; Engineering, Civil; Materials Science, Multidisciplinary', 'Rehabilitation; Sport Sciences', 'Engineering, Mechanical', 'Engineering, Marine; Engineering, Civil; Engineering, Ocean; Oceanography', 'Engineering, Multidisciplinary', 'Neuroimaging', 'Pediatrics', 'Automation & Control Systems; Operations Research & Management Science', 'Cardiac & Cardiovascular Systems; Pathology', 'Behavioral Sciences; Neurosciences; Psychology, Experimental', 'Emergency Medicine', 'Nanoscience & Nanotechnology; Materials Science, Multidisciplinary; Metallurgy & Metallurgical Engineering', 'Virology', 'Archaeology', 'Marine & Freshwater Biology', 'Obstetrics & Gynecology; Reproductive Biology', 'Geosciences, Multidisciplinary', 'Management', 'Endocrinology & Metabolism; Neurosciences; Psychiatry', 'Biology; Computer Science, Interdisciplinary Applications; Engineering, Biomedical; Mathematical & Computational Biology', 'Medicine, General & Internal; Medicine, Research & Experimental', 'Information Science & Library Science', 'Ecology; Parasitology', 'Engineering, Civil', 'Engineering, Industrial; Operations Research & Management Science', 'Critical Care Medicine; Emergency Medicine', 'Nanoscience & Nanotechnology; Materials Science, Multidisciplinary', 'Biochemical Research Methods', 'Ergonomics; Public, Environmental & Occupational Health; Social Sciences, Interdisciplinary; Transportation', 'Engineering, Mechanical; Materials Science, Multidisciplinary', 'Computer Science, Information Systems; Telecommunications', 'Hematology; Immunology; Transplantation', 'Genetics & Heredity', 'Materials Science, Multidisciplinary; Physics, Applied', 'Astronomy & Astrophysics; Physics, Nuclear; Physics, Particles & Fields', 'Ecology; Economics; Environmental Sciences; Environmental Studies', 'Environmental Studies; Transportation; Transportation Science & Technology', 'Education & Educational Research', 'Rehabilitation', 'Biotechnology & Applied Microbiology; Chemistry, Analytical; Electrochemistry; Nanoscience & Nanotechnology; Instruments & Instrumentation', 'Engineering, Environmental; Engineering, Chemical', 'Public, Environmental & Occupational Health; Social Sciences, Biomedical', 'Pathology', 'Reproductive Biology; Veterinary Sciences', 'Toxicology', 'Immunology', 'Chemistry, Applied; Food Science & Technology', 'Public, Environmental & Occupational Health; Pharmacology & Pharmacy', 'Entomology', 'Education & Educational Research; Psychology, Educational', 'Computer Science, Artificial Intelligence; Computer Science, Interdisciplinary Applications', 'Obstetrics & Gynecology; Peripheral Vascular Disease', 'Biochemistry & Molecular Biology; Biophysics', 'Construction & Building Technology; Engineering, Environmental; Engineering, Civil', 'Neurosciences; Physiology; Rehabilitation; Sport Sciences', 'Clinical Neurology; Psychiatry', 'Psychology, Social', 'Geography', 'Biotechnology & Applied Microbiology; Genetics & Heredity; Toxicology', 'Green & Sustainable Science & Technology; Environmental Sciences', 'Psychology, Biological; Behavioral Sciences; Psychology; Psychology, Experimental', 'Automation & Control Systems; Chemistry, Analytical; Computer Science, Artificial Intelligence; Instruments & Instrumentation; Mathematics, Interdisciplinary Applications; Statistics & Probability', 'Engineering, Aerospace', 'Food Science & Technology; Microbiology', 'Clinical Neurology; Neurosciences', 'Computer Science, Artificial Intelligence; Engineering, Electrical & Electronic', 'Agriculture, Dairy & Animal Science; Veterinary Sciences', 'Veterinary Sciences', 'Otorhinolaryngology', 'Ecology; Environmental Sciences; Environmental Studies', 'Engineering, Manufacturing', 'Social Sciences, Interdisciplinary', 'Integrative & Complementary Medicine', 'Computer Science, Interdisciplinary Applications; Computer Science, Theory & Methods', 'Economics; Transportation', 'Oceanography', 'Marine & Freshwater Biology; Oceanography', 'Biochemistry & Molecular Biology; Chemistry, Organic', 'Neurosciences; Pharmacology & Pharmacy', 'Computer Science, Artificial Intelligence; Engineering, Multidisciplinary', 'Chemistry, Physical; Materials Science, Multidisciplinary', 'Materials Science, Multidisciplinary; Nuclear Science & Technology', 'Engineering, Electrical & Electronic; Materials Science, Multidisciplinary; Physics, Applied; Physics, Condensed Matter', 'Peripheral Vascular Disease', 'Medicine, Legal; Pharmacology & Pharmacy; Toxicology', 'Agronomy; Forestry; Meteorology & Atmospheric Sciences', 'Computer Science, Interdisciplinary Applications; Engineering, Industrial; Operations Research & Management Science', 'Economics; Energy & Fuels; Environmental Sciences; Environmental Studies', 'Oncology', 'Biochemistry & Molecular Biology; Entomology', 'Chemistry, Physical; Nanoscience & Nanotechnology; Materials Science, Multidisciplinary', 'Engineering, Mechanical; Mechanics', 'Microbiology', 'Geosciences, Multidisciplinary; Meteorology & Atmospheric Sciences; Water Resources', 'Ecology; Mycology', 'Computer Science, Information Systems', 'Agronomy; Food Science & Technology; Horticulture', 'Oncology; Public, Environmental & Occupational Health', 'Biochemical Research Methods; Biotechnology & Applied Microbiology; Virology', 'Acoustics', 'Horticulture', 'Green & Sustainable Science & Technology; Geography, Physical', 'Biotechnology & Applied Microbiology', 'Chemistry, Applied; Chemistry, Physical; Nanoscience & Nanotechnology; Materials Science, Multidisciplinary', 'Business; Communication', 'Engineering, Chemical; Food Science & Technology', 'Neurosciences; Pharmacology & Pharmacy; Toxicology', 'Respiratory System', 'Oncology; Cardiac & Cardiovascular Systems', 'Parasitology; Pharmacology & Pharmacy', 'Microscopy', 'History & Philosophy Of Science', 'Anthropology; Archaeology; Geosciences, Multidisciplinary', 'Environmental Sciences; Remote Sensing; Imaging Science & Photographic Technology', 'Orthopedics; Sport Sciences; Surgery', 'Agriculture, Dairy & Animal Science; Behavioral Sciences; Veterinary Sciences', 'Family Studies; Psychology, Social; Social Work', 'Chemistry, Analytical; Electrochemistry', 'Business; Management', 'Hospitality, Leisure, Sport & Tourism; Psychology, Applied; Psychology; Sport Sciences', 'Engineering, Aerospace; Astronomy & Astrophysics; Geosciences, Multidisciplinary; Meteorology & Atmospheric Sciences', 'Mathematics', 'Oncology; Respiratory System', 'Computer Science, Interdisciplinary Applications; Computer Science, Theory & Methods; Engineering, Biomedical; Medical Informatics', 'Engineering, Manufacturing; Materials Science, Multidisciplinary', 'Anatomy & Morphology', 'Biochemistry & Molecular Biology; Chemistry, Applied; Polymer Science', 'Computer Science, Interdisciplinary Applications; Physics, Mathematical', 'Computer Science, Theory & Methods; Mathematics, Applied', 'Astronomy & Astrophysics', 'Instruments & Instrumentation; Optics; Physics, Applied', 'Primary Health Care; Medicine, General & Internal', 'Geriatrics & Gerontology; Neurosciences', 'Materials Science, Multidisciplinary; Physics, Condensed Matter', 'Psychology, Clinical; Health Care Sciences & Services; Medical Informatics; Psychiatry', 'Geosciences, Multidisciplinary; Oceanography', 'Agricultural Engineering; Agriculture, Multidisciplinary', 'Engineering, Multidisciplinary; Instruments & Instrumentation', 'Cell Biology', 'Biochemistry & Molecular Biology; Cell Biology', 'Computer Science, Interdisciplinary Applications; Medical Informatics', 'Fisheries; Marine & Freshwater Biology', 'Agronomy', 'Immunology; Pharmacology & Pharmacy', 'Construction & Building Technology; Materials Science, Composites', 'Chemistry, Medicinal; Pharmacology & Pharmacy', 'Chemistry, Applied; Materials Science, Coatings & Films', 'Geography; Political Science', 'Computer Science, Interdisciplinary Applications; Engineering, Environmental; Environmental Studies; Geography; Operations Research & Management Science; Regional & Urban Planning', 'Oncology; Obstetrics & Gynecology', 'Substance Abuse', 'Materials Science, Multidisciplinary; Metallurgy & Metallurgical Engineering; Materials Science, Characterization & Testing', 'Biotechnology & Applied Microbiology; Food Science & Technology; Microbiology', 'Anesthesiology', 'Chemistry, Physical; Engineering, Environmental; Engineering, Chemical', 'Chemistry, Physical; Nuclear Science & Technology; Physics, Atomic, Molecular & Chemical', 'Agriculture, Dairy & Animal Science; Food Science & Technology', 'Mathematics, Applied; Mathematics', 'Medicine, Research & Experimental', 'Mathematics, Interdisciplinary Applications; Physics, Multidisciplinary; Physics, Mathematical', 'Environmental Studies; Psychology, Multidisciplinary', 'Electrochemistry', 'Ecology', 'Environmental Sciences; Environmental Studies; Geography', 'Polymer Science', 'Immunology; Microbiology; Respiratory System', 'Biochemistry & Molecular Biology; Pharmacology & Pharmacy; Toxicology', 'Chemistry, Medicinal', 'Construction & Building Technology; Green & Sustainable Science & Technology; Energy & Fuels', 'Economics; Law', 'Education & Educational Research; Linguistics; Language & Linguistics', 'Economics; Public, Environmental & Occupational Health', 'Biochemical Research Methods; Chemistry, Analytical', 'Materials Science, Multidisciplinary; Mechanics; Physics, Condensed Matter', 'Automation & Control Systems; Engineering, Electrical & Electronic; Engineering, Mechanical; Robotics', 'Thermodynamics; Energy & Fuels; Engineering, Mechanical; Mechanics', 'Materials Science, Composites', 'Social Issues; Social Sciences, Interdisciplinary', 'Biotechnology & Applied Microbiology; Microbiology', 'Computer Science, Interdisciplinary Applications; Engineering, Environmental; Environmental Sciences; Water Resources', 'Biochemical Research Methods; Microbiology', 'Medicine, General & Internal; Urology & Nephrology', 'Behavioral Sciences; Clinical Neurology; Psychiatry', 'Chemistry, Physical; Physics, Condensed Matter', 'Chemistry, Medicinal; Chemistry, Multidisciplinary; Pharmacology & Pharmacy', 'Chemistry, Applied; Chemistry, Organic; Polymer Science', 'Biology', 'Environmental Sciences; Limnology; Marine & Freshwater Biology', 'Endocrinology & Metabolism; Primary Health Care', 'Chemistry, Physical; Electrochemistry; Energy & Fuels', 'Engineering, Electrical & Electronic; Instruments & Instrumentation; Materials Science, Multidisciplinary', 'Engineering, Industrial; Engineering, Manufacturing; Materials Science, Multidisciplinary', 'Medical Laboratory Technology', 'Cell Biology; Geriatrics & Gerontology', 'Biotechnology & Applied Microbiology; Genetics & Heredity', 'Biochemical Research Methods; Biochemistry & Molecular Biology', 'Chemistry, Physical; Electrochemistry; Energy & Fuels; Materials Science, Multidisciplinary', 'Acoustics; Engineering, Biomedical; Instruments & Instrumentation; Radiology, Nuclear Medicine & Medical Imaging', 'Computer Science, Artificial Intelligence; Computer Science, Information Systems; Computer Science, Software Engineering', 'Biochemistry & Molecular Biology; Parasitology', 'Thermodynamics; Chemistry, Physical; Engineering, Chemical', 'Orthopedics; Rheumatology', 'Biochemistry & Molecular Biology; Endocrinology & Metabolism', 'Geosciences, Multidisciplinary; Soil Science; Water Resources', 'Thermodynamics; Engineering, Mechanical', 'Business; Management; Transportation', 'Education & Educational Research; Linguistics', 'Psychology, Multidisciplinary; Psychology, Experimental', 'Computer Science, Interdisciplinary Applications', 'Computer Science, Theory & Methods; Logic', 'Engineering, Geological; Mining & Mineral Processing', 'Mathematics, Applied; Mathematics; Physics, Mathematical', 'Developmental Biology', 'Psychology, Social; Social Sciences, Interdisciplinary; Sociology', 'Mathematical & Computational Biology; Infectious Diseases', 'Biotechnology & Applied Microbiology; Engineering, Environmental; Environmental Sciences', 'Food Science & Technology; Toxicology', 'Acoustics; Computer Science, Interdisciplinary Applications', 'Engineering, Civil; Geosciences, Multidisciplinary; Water Resources', 'Sociology', 'Materials Science, Ceramics', 'Biophysics; Biotechnology & Applied Microbiology; Chemistry, Analytical; Electrochemistry; Nanoscience & Nanotechnology', 'Biochemistry & Molecular Biology; Evolutionary Biology; Genetics & Heredity', 'Thermodynamics; Energy & Fuels; Mechanics', 'Engineering, Environmental; Environmental Sciences; Water Resources', 'Rheumatology', 'Cardiac & Cardiovascular Systems; Pediatrics', 'Public, Environmental & Occupational Health; Medicine, General & Internal', 'Psychology, Developmental', 'Energy & Fuels; Engineering, Petroleum', 'Clinical Neurology; Neuroimaging; Psychiatry', 'Environmental Sciences; Pharmacology & Pharmacy; Toxicology', 'Chemistry, Applied; Energy & Fuels; Engineering, Chemical', 'Oncology; Cell Biology', 'Biochemical Research Methods; Biotechnology & Applied Microbiology; Chemistry, Analytical', 'Infectious Diseases; Microbiology; Pharmacology & Pharmacy', 'Engineering, Biomedical', 'Biochemistry & Molecular Biology; Plant Sciences', 'Obstetrics & Gynecology; Pediatrics', 'Automation & Control Systems; Engineering, Multidisciplinary; Instruments & Instrumentation', 'Economics; Management', 'Computer Science, Interdisciplinary Applications; Engineering, Geological; Geosciences, Multidisciplinary', 'Mycology', 'Materials Science, Multidisciplinary; Materials Science, Coatings & Films; Physics, Applied; Physics, Condensed Matter', 'Operations Research & Management Science', 'Operations Research & Management Science; Transportation Science & Technology', 'Fisheries; Immunology; Marine & Freshwater Biology; Veterinary Sciences', 'Immunology; Neurosciences', 'Plant Sciences', 'Political Science', 'Agronomy; Water Resources', 'Hospitality, Leisure, Sport & Tourism', 'Oncology; Hematology', 'Biochemistry & Molecular Biology; Cell Biology; Immunology', 'Information Science & Library Science; Management', 'Engineering, Industrial; Ergonomics', 'Engineering, Environmental; Engineering, Civil; Geosciences, Multidisciplinary', 'Agricultural Economics & Policy; Economics; Food Science & Technology; Nutrition & Dietetics', 'Biochemistry & Molecular Biology; Nutrition & Dietetics', 'Immunology; Microbiology', 'Literature; Sociology', 'Parasitology; Veterinary Sciences', 'Agriculture, Multidisciplinary; Ecology; Environmental Sciences', 'Business; Economics; Environmental Studies', 'Engineering, Civil; Engineering, Ocean', 'Cell Biology; Endocrinology & Metabolism', 'Biochemistry & Molecular Biology; Neurosciences', 'Agriculture, Dairy & Animal Science; Endocrinology & Metabolism', 'Computer Science, Artificial Intelligence; Engineering, Biomedical; Medical Informatics', 'Biophysics; Engineering, Biomedical', 'Chemistry, Applied; Engineering, Chemical; Materials Science, Textiles', 'Computer Science, Information Systems; Engineering, Electrical & Electronic; Telecommunications', 'Audiology & Speech-Language Pathology; Otorhinolaryngology', 'Environmental Sciences; Nanoscience & Nanotechnology', 'Cardiac & Cardiovascular Systems; Cell Biology', 'Chemistry, Multidisciplinary; Materials Science, Multidisciplinary; Polymer Science', 'Engineering, Chemical; Materials Science, Multidisciplinary', 'Optics; Spectroscopy', 'Health Policy & Services', 'Engineering, Electrical & Electronic', 'Microbiology; Veterinary Sciences', 'Environmental Sciences; Food Science & Technology; Microbiology', 'Geography, Physical; Geosciences, Multidisciplinary; Paleontology', 'Biochemical Research Methods; Biotechnology & Applied Microbiology', 'Engineering, Chemical; Mineralogy; Mining & Mineral Processing', 'Neurosciences; Ophthalmology; Psychology', 'Agricultural Engineering; Fisheries', 'Astronomy & Astrophysics; Physics, Particles & Fields', 'Geriatrics & Gerontology', 'Economics; Geography; Transportation', 'Critical Care Medicine', 'Orthopedics; Sport Sciences', 'Biodiversity Conservation; Ecology; Environmental Sciences', 'Engineering, Multidisciplinary; Materials Science, Composites', 'Oceanography; Water Resources', 'Biodiversity Conservation; Ecology', 'Criminology & Penology; Psychology, Multidisciplinary', 'Psychology, Clinical; Psychiatry', 'Biotechnology & Applied Microbiology; Engineering, Chemical; Food Science & Technology', 'Engineering, Multidisciplinary; Engineering, Manufacturing', 'Engineering, Manufacturing; Materials Science, Composites', 'Computer Science, Information Systems; Information Science & Library Science', 'Energy & Fuels; Environmental Sciences; Environmental Studies', 'Plant Sciences; Environmental Studies; Forestry; Urban Studies', 'Dermatology', 'Education, Scientific Disciplines; Nutrition & Dietetics', 'Thermodynamics; Chemistry, Analytical; Chemistry, Physical', 'Materials Science, Multidisciplinary; Multidisciplinary Sciences', 'Biochemical Research Methods; Physics, Atomic, Molecular & Chemical; Spectroscopy', 'Engineering, Industrial; Engineering, Manufacturing; Operations Research & Management Science', 'Plant Sciences; Environmental Sciences', 'Computer Science, Cybernetics; Ergonomics; Psychology, Multidisciplinary', 'Psychology, Applied; Transportation', 'Economics; Engineering, Civil; Operations Research & Management Science; Transportation; Transportation Science & Technology', 'Engineering, Geological; Geosciences, Multidisciplinary', 'Management; Operations Research & Management Science', 'Education, Special; Rehabilitation', 'Computer Science, Hardware & Architecture; Computer Science, Theory & Methods', 'Biochemistry & Molecular Biology; Biology; Biophysics; Electrochemistry', 'Medicine, Research & Experimental; Clinical Neurology', 'Automation & Control Systems; Engineering, Electrical & Electronic', 'Cardiac & Cardiovascular Systems; Endocrinology & Metabolism; Nutrition & Dietetics', 'Biochemistry & Molecular Biology; Genetics & Heredity', 'Public, Environmental & Occupational Health; Transportation', 'Linguistics; Language & Linguistics', 'Plant Sciences; Ecology', 'Cardiac & Cardiovascular Systems; Radiology, Nuclear Medicine & Medical Imaging', 'Green & Sustainable Science & Technology; Energy & Fuels; Engineering, Environmental; Engineering, Chemical', 'Reproductive Biology; Toxicology', 'Sport Sciences', 'Acoustics; Engineering, Mechanical; Mechanics', 'Materials Science, Ceramics; Materials Science, Multidisciplinary', 'Engineering, Chemical; Water Resources', 'Chemistry, Inorganic & Nuclear', 'Computer Science, Hardware & Architecture; Computer Science, Information Systems; Engineering, Electrical & Electronic; Telecommunications', 'Critical Care Medicine; Emergency Medicine; Orthopedics; Surgery', 'Gerontology', 'Computer Science, Artificial Intelligence; Computer Science, Interdisciplinary Applications; Engineering, Biomedical; Radiology, Nuclear Medicine & Medical Imaging', 'Environmental Sciences; Environmental Studies; Geosciences, Multidisciplinary; Public, Environmental & Occupational Health; Meteorology & Atmospheric Sciences', 'Economics; Political Science', 'Fisheries; Immunology; Veterinary Sciences; Zoology', 'Audiology & Speech-Language Pathology; Neurosciences; Otorhinolaryngology', 'Chemistry, Analytical; Energy & Fuels; Engineering, Chemical', 'Development Studies; Economics', 'Geosciences, Multidisciplinary; Mining & Mineral Processing', 'Chemistry, Physical; Energy & Fuels; Materials Science, Multidisciplinary', 'Biochemistry & Molecular Biology; Endocrinology & Metabolism; Toxicology; Zoology', 'Psychology, Biological; Neurosciences; Physiology; Psychology; Psychology, Experimental', 'Biochemical Research Methods; Immunology', 'Oncology; Dentistry, Oral Surgery & Medicine', 'Engineering, Marine', 'Chemistry, Physical; Nanoscience & Nanotechnology; Materials Science, Multidisciplinary; Physics, Applied', 'Health Care Sciences & Services; Health Policy & Services', 'Oncology; Geriatrics & Gerontology', 'Medicine, Legal', 'Ecology; Environmental Studies; Geography; Geography, Physical; Regional & Urban Planning; Urban Studies', 'Biochemistry & Molecular Biology; Chemistry, Medicinal; Chemistry, Organic', 'Chemistry, Physical; Materials Science, Multidisciplinary; Mineralogy', 'Physiology; Respiratory System', 'Critical Care Medicine; Respiratory System', 'Psychology, Biological; Behavioral Sciences; Social Sciences, Biomedical', 'Genetics & Heredity; Medicine, Legal', 'Behavioral Sciences; Neurosciences; Pharmacology & Pharmacy', 'Astronomy & Astrophysics; Biology; Multidisciplinary Sciences', 'Oncology; Nursing', 'Law', 'Ecology; Marine & Freshwater Biology', 'Cell Biology; Immunology', 'Anesthesiology; Cardiac & Cardiovascular Systems; Respiratory System; Peripheral Vascular Disease', 'Orthopedics; Surgery', 'Construction & Building Technology; Materials Science, Multidisciplinary', 'Psychology, Applied', 'Thermodynamics; Energy & Fuels; Engineering, Chemical; Engineering, Mechanical', 'Materials Science, Multidisciplinary; Mechanics', 'History & Philosophy Of Science; Mathematics, Interdisciplinary Applications; Mathematics', 'Biodiversity Conservation; Biology', 'Biotechnology & Applied Microbiology; Food Science & Technology', 'Computer Science, Artificial Intelligence; Energy & Fuels', 'Education, Scientific Disciplines; Nursing', 'Business, Finance; Economics; Management; Operations Research & Management Science', 'Cardiac & Cardiovascular Systems; Respiratory System', 'Immunology; Infectious Diseases; Virology', 'Environmental Sciences; Soil Science; Water Resources', 'Genetics & Heredity; Toxicology', 'Remote Sensing', 'Energy & Fuels; Engineering, Chemical; Engineering, Petroleum', 'Transportation; Transportation Science & Technology', 'Neurosciences; Psychology; Psychology, Experimental; Sport Sciences', 'Computer Science, Information Systems; Computer Science, Interdisciplinary Applications', 'Biochemistry & Molecular Biology; Mathematical & Computational Biology', 'Infectious Diseases; Virology', 'Engineering, Multidisciplinary; Mechanics', 'Computer Science, Interdisciplinary Applications; Geosciences, Multidisciplinary', 'Health Care Sciences & Services; Health Policy & Services; Public, Environmental & Occupational Health', 'Engineering, Mechanical; Materials Science, Multidisciplinary; Mechanics', 'Energy & Fuels; Engineering, Petroleum; Engineering, Mechanical', 'Economics; Environmental Studies; Forestry', 'Endocrinology & Metabolism; Orthopedics', 'Endocrinology & Metabolism; Nutrition & Dietetics', 'Developmental Biology; Obstetrics & Gynecology; Reproductive Biology', 'Chemistry, Inorganic & Nuclear; Crystallography', 'Dentistry, Oral Surgery & Medicine; Surgery', 'Psychology, Developmental; Psychology, Experimental', 'Thermodynamics; Energy & Fuels; Engineering, Multidisciplinary; Engineering, Chemical; Engineering, Mechanical', 'Instruments & Instrumentation; Nuclear Science & Technology; Physics, Atomic, Molecular & Chemical; Physics, Nuclear','Other']

selected_category = st.selectbox("Select WoS categories:", options, index= None)
if selected_category == "Other":
    custom_category = st.text_input("Enter custom category:")
    selected_category = custom_category if custom_category else "Other"

combined_text = f"{title_input} [SEP] {keywords_input} [SEP] {abstract_input} [SEP] {selected_category} [SEP] {affiliations_input}"
if st.button("Predict"):
    if not any([title_input, abstract_input,keywords_input,  affiliations_input]):
        st.warning("Please enter paper text.")
    else:
        with st.spinner("Predicting..."):
            predicted_class = predict_class(combined_text)
            if predicted_class is not None:
                class_labels = ["Level 1", "Level 2", "Level 3", "Level 4"]

                st.text("Predicted Class:")
                for i, label in enumerate(class_labels):
                    if i == predicted_class:
                        st.markdown(
                            f'<div style="background-color: {class_colors[predicted_class]}; padding: 10px; border-radius: 5px; color: white; font-weight: bold;">{label}</div>',
                            unsafe_allow_html=True
                        )
                    else:
                        st.text(label)

#elif option == "PDF":
#    uploaded_file = st.file_uploader("Upload a PDF file", type=["pdf"])

#    if uploaded_file is not None:
#        with st.spinner("Processing PDF..."):
#            file_path = os.path.join(uploaded_files_dir, uploaded_file.name)
#            with open(file_path, "wb") as f:
#                f.write(uploaded_file.getbuffer())
#            st.success("File uploaded successfully.")
#            st.text(f"File Path: {file_path}")
#            
#            file_text = extract_text_from_pdf(file_path)
#            st.text("Extracted Text:")
#            st.text(file_text)
#
#            if st.button("Predict from PDF Text"):
#                if not file_text.strip():
#                    st.warning("Please upload a PDF with text content.")
#                else:
#                    with st.spinner("Predicting..."):
#                        predicted_class = predict_class(file_text)
#                        if predicted_class is not None:
#                            class_labels = ["Level 1", "Level 2", "Level 3", "Level 4"]
#                            st.text("**Predicted Class:**")
#                            for i, label in enumerate(class_labels):
#                                if i == predicted_class:
#                                    st.markdown(
#                                        f'<div style="background-color: {class_colors[predicted_class]}; padding: 10px; border-radius: 5px; color: white; font-weight: bold;">{label}</div>',
#                                        unsafe_allow_html=True
#                                    )
#                                else:
#                                    st.text(label)