Senior_Capstone

Team Name

JTCX

Team Memebers

Jordan Shaheen

Major: Computer Science
Email Address: shaheejn@mail.uc.edu

Cole Hutchins

Major: Computer Science
Email Address: hutchicj@mail.uc.edu

Toby Knueven

Major: Computer Science
Email Address: knueveta@mail.uc.edu

Xander Hatton

Major: Computer Science
Email Address: hattonab@mail.uc.edu

Dr. William Hawkins III

Advisor
hawkinwh@ucmail.uc.edu

Project Abstract

The News Bias Detector aims to promote media literacy by analyzing online news for bias. Utilizing advanced NLP and ML techniques, it scrutinizes text for language patterns indicating bias. This project, driven by the escalating concerns around misinformation, leverages Django for web development and Python for algorithmic analysis. Key features include URL input for scraping articles and real-time bias assessment. Development is informed by extensive research into bias detection methodologies, existing tools, and user interaction paradigms, ensuring a blend of technical proficiency and user-centric design. This tool aspires to be an essential aid for discerning readers in the digital age.

User Stories

As a person reading the news, I want to understand the bias of articles I read in order to come to my own understanding of topics I am interested in.
As a Polysci student, I want a bias-metric to understand the mathmatical measurement of the bias in order to research more effectively.
As a marketing professional, I want to understand the bias in news articles to help advertise correctly to the right audiences.

Design Diagrams

Design Level 0

graph LR;
    A[User] --> B;
    B[Send News Article to Website] --> C;
    C[Obtain Bias Rating of the Article];

Design Level 1

graph TD;
    A[User] --> B;
    B[Access Bias News Detector Website] --> C;
    C[Upload a News Website's Article Link] --> D;
    D[Scrape Text from Article Link] --> E;
    E[Determine Bias Rating of the Article];

Design Level 2

graph TD;
    A[User] --> B;
    B[Access Bias News Detector Website] --> C;
    C[Upload a News Website's Article Link] --> D;
    D[Scrape Text from Article Link] --> E;
    E[Use Natural Language Processing to Find Bias Words] --> F;
    F[Measure Bias Score Based on the Context, Author, Information Given, etc.] --> G;
    G[Reveal Bias Metric to User and Highlighted Bias Sections];

Project Task List

Create a formalized outline of webpage mapping and universal styling for all webpages
Create webscraper that analyizes the news article website URL and scrapes the article text accuratly
Create Machine Learning Natural Language Processing feature that can correctly detect bias in the scraped news article
Create bias metric with Machine Learning to understand how bias or unbias the news article is and present the metric intuitivly to user.

Project Timeline

Task	Start Date	Planned Completion Date
Task 1: Webpage outline	01/08/2024	01/22/2024
Task 2: Webscraper	01/22/2024	02/05/2024
Task 3: Bias metric/training data	02/05/2024	02/19/2024
Task 4: ML NLP bias classifier	02/19/2024	03/18/2024

Effort Matrix

Task Description	Team Member of Primary Responsibility	Shaheen Effort (%)	Hatton Effort (%)	Knueven Effort (%)	Hutchins Effort (%)
Task 1: Webpage outline	Jordan Shaheen	75%	25%	0%	0%
Task 2: Webscraper	Xander Hatton	25%	75%	0%	0%
Task 3: Bias metric/training data	Tobias Knueven	0%	0%	75%	25%
Task 4: ML NLP bias classifier	Colson Hutchins	0%	0%	25%	75%

ABET Concerns Essay

The development and deployment of our news bias detector are influenced by multiple economic, ethical, security, and social constraints. Economically, our project relies heavily on open-source software and publicly available tools due to the short time for development, and our team has no budget for the project. These circumstances may limit the range of features we can offer and the quality of our bias detection. Professionally, this project aims to bolster the reputation of an unbiased, factual news distribution. From an ethical standpoint, our detector's potential influence on a user's perception of news presents a significant challenge. Our tool mustn't inadvertently promote a particular narrative or suppress another, ensuring its impact remains neutral. Security concerns also arise, mainly related to user data privacy. Since users will submit news article URLs for analysis, measures to prevent data breaches and ensure the anonymity of user submissions are paramount. Socially, the news bias detector is intended to serve the broader public by promoting informed citizenship. By helping users discern biases in news, we aspire to enhance the quality of public discourse. Environmental and cultural constraints are less predominant in our project; however, our design is focused on accessibility, ensuring diverse user groups, irrespective of language or background, can benefit from our tool.