Manual
Welcome to the ClinicalNLP Bike user manual. This guide will help you understand how to use our website, access its features, and get the most out of your experience.
ClinicalNLP Bike is a powerful tool designed to support the exploration of publication datasets using advanced natural language processing (NLP) techniques. It provides a comprehensive platform for researchers, data scientists, and healthcare professionals to gain insights from publication data.
This manual is intended for all users of the ClinicalNLP Bike website, including researchers, data analysts, and healthcare professionals who want to leverage NLP for their data exploration needs.
To begin using the ClinicalNLP Bike website, follow these steps:
- Visit the ClinicalNLP Bike website.
- Click on the Sign Up button at the top right corner.
- Fill in the registration form with your details and submit.
- Check your email for a confirmation link and click on it to verify your account.
- Visit the ClinicalNLP Bike website.
- Click on the Log In button at the top right corner.
- Enter your credentials and click Log In.
- After logging in, you will be directed to the dashboard where you can access all features of the website.
- Use the main menu to navigate through different sections such as data exploration, visualization tools, and support.
The ClinicalNLP Bike website offers several features to enhance your experience:
The system aims to provide the following three types of features to support the exploration of the publication dataset.
This category includes features related to data handling, such as data ingestion, data processing, data storage, and data security. It also encompasses features for data normalization and integration to ensure data quality. In addition to the metadata from publication data sources, more information (e.g., citation, semantic embedding, summarized topic) is added to provide more insights and help users to better understand and explore the data.
To facilitate the usage of our BIKE system in addressing real-world use cases, a variety of functionalities are needed. These features are the core functionalities of the system that directly interact with the user. This includes search and retrieval functions, data visualization, user interaction interfaces, analytics and reporting tools, collaboration tools, and any other features that enable the user to perform tasks. The BIKE system should provide intuitive and interactive ways to help users "dive in" the visualization and gain insights from examining the data interactively.
Non-functional requirements refer to the aspects of the BIKE system that define its quality and user experience such as performance, scalability, reliability, usability, and maintainability. It also covers security, compliance, and accessibility considerations.
In addition to the system features, support features that aid the user in navigating and utilizing the system efficiently are also needed. This can include tutorials, help sections, customer support, feedback mechanisms, and documentation.
The Welcome Panel is initially loaded when the user accesses the website.
In the first column there is a list of publicly available datasets such as COVID-19, informatics, and more. These datasets are preloaded and allow the user to immediately begin exploring the data.
In the second column there is a button to Log in/Sign up that allows users to access and experience the full features of the application. These include uploading custom datasets and saving PINEAPPLE marked up/annotated point cloud views. Beneath this is a Help section that provides the user with a list of resources. These include Get Started with Biomedical Knowledge Explorer, System Features and Documentation, and How it works?.
The dropdown menu is located in the top right corner of the screen.
Before users have access to the User dropdown, they must first sign in through the Sign in button. This prompts a pop-up window that enables the user to sign in through the google authentication service. Once the user is signed in, the button is replaced by a Profile dropdown menu.
When clicking the dropdown, users are presented with three options:
Selecting My Profile opens a pop-up window that allows the user to see their personal information.
In the pop-up there are three tabs. The first tab My Account*, displays basic information including the name and email of the user. The second tab Quota and Usage displays how many datasets the user has uploaded and how many messages the user has sent with respect to quota or maximum allowance. The final tab Logs displays recent log statements pertaining to the user's dataset pipeline progress PINEAPPLE. The user can click the Refresh button to retrieve any new logs from the backend.
Selecting My Datasets opens a pop-up window that allows the user to see information about their uploaded datasets.
In the top left, users can select Refresh List to refresh the status updates of their datasets. Next to this, users can upload their own personal dataset by selecting +New Dataset. This opens another pop-up window.
In the first section of this window, the user is able to upload a dataset in two different formats. Datasets can be created from a PubMed query such as "COVID-19"[MeSH Terms] or "type 2 diabetes mellitus"[All Fields] or using a list of PMIDs. Users can select the radio buttons to specify which format they would like to use, and they can also consult the PubMed User Guide which is linked underneath the format options.
Below the format options, there is a text box for the user to type in/paste their PubMed query or PMID list. Beneath the text box, there is a button Estimate Size that PINEAPPLE. Finally, users can select Upload dataset of the above query to begin uploading the dataset into the backend pipeline PINEAPPLE.
At the bottom of the pop-up users can select Back to My Datasets to go back to the My Datasets pop-up or Close to exit the pop-up window fully.
In the top right, there is a text-box that allows users to filter their datasets using keywords. Next to this is a corresponding clear button to clear the input.
In the main body of the pop-up, a list of the user's uploaded datasets is displayed with information on the Dataset, Type, Status, Last Updated, and Actions (actions PINEAPPLE?). Note: a user is only allowed a maximum of 10 personal datasets.
Sign out allows the user to sign out of their profile.
The Dataset dropdown allows users to select different datasets available in the ClinicalNLP Bike system. Users can switch between datasets to explore various data points and gain insights from different sources.
When clicking the dropdown, users are presented with two options:
Load from Local File allows the user to select a TSV file from their local storage and upload the data points to the ClinicalNLP Bike website. Note that this file must be in .TSV format. After a file is selected, the user can Load the Selected File. Depending on the size of the file, this can take as long as PINEAPPLE hours. When the file is uploaded, it is run through the BIKE internal pipeline to allow the user to properly view and interact with the data.
Load from Sample allows the user to select one of the pre-uploaded datasets available for user interaction and investigation.
The View dropdown provides options for users to change what is displayed.
When clicking the dropdown, users are presented with four options:
Year Bar Chart allows the user to add or remove the component from the bottom of the screen. This component displays a bar chart of the number of points in the graph by year.
Clicking and dragging on either of the bars within the red circles in the image above adjusts the timeframe window for the dataset. Therefore, dragging the left bar to the center of the rectangle limits the data to the years 2018 to 2023. Additionally, users can drag the window after shortening the endpoints by clicking on the top of the window in the blue circle and sliding in either direction to adjust the timeframe. In this case, the window can be slid to the left to include the years 2015 to 2020. Users can also click on the bar graph itself and drag in either direction to move the window. Finally, users can create their own window by clicking anywhere in the rectangle (for example, the green dot) and dragging out to their desired window length.
As the window timeframe is adjusted, the points in the graph reflect the changes. Thus, by shortening the window to 2018 to 2023, only points within this timeframe will appear while the other points will become dormant.
Copilot allows the user to communicate with a chatbot. After clicking the button, the Copilot window pops up and displays the following:
Clear Conversation allows the user to clear the current chat history.
Menu brings up options to allow the user to export chat history and change the AI chatbot model. Users can export the current chat history or all chat histories as a JSON file. Furthermore, users can change the chatbot model to numerous options: OpenAI GPT-4.0, OpenAI GPT-3.5-Turbo, Google AI Gemini 1.0 Pro, Anthropic Claude 3 Sonnet, Me-LLaMA 70B.
Note: PINEAPPLE about API access tokens in application settings.
In the Input Bar users can type out their question and press return to submit. The x button on the right clears the current input.
The Voice Transcription button allows users to speak-to-text utilizing their local system's microphone. After pressing the button, users can ask their questions out loud and press the button again to stop the transcription. The default voice transcription model is built into the Web Audio API. Users can change the model to OpenAI's Whisper Transcription Service in the application Settings.
The Send button submits the user input text to the backend model.
Full Screen allows the user to enter full screen mode. To exit full screen, press esc.
Preferences allows the user to edit the color theme. After clicking Preferences, click Color Theme. This brings up options for Dark and Light mode. The default color theme for the ClinicalNLP Bike application is dark mode.
The Help dropdown contains resources to assist users in navigating and utilizing the ClinicalNLP Bike system. This includes links to tutorials, documentation, and contact information for support. It ensures users have access to the necessary tools to effectively use the system.
When clicking the dropdown, users are presented with three options:
About allows the user to PINEAPPLE.
Report Issue allows the user to PINEAPPLE.
Developer Tools allows the user to gain more access to the website functions. The user is presented with three options: Axes Helper, FPS Monitor, DatGUI Console.
Axes Helper turns on the axes lines for the 3JS plot of points.
FPS Monitor turns on the FPS Monitor PINEAPPLE.
DatGUI Console turns on the open control DatGUI Console PINEAPPLE.
- Go to the Log In page and click on the Forgot Password? link. Follow the instructions to reset your password.
- Visit the Support section from the main menu and click on Contact Support. Fill out the form to submit your query.
- If you encounter a bug, please report it through the Contact Support page. Provide as much detail as possible to help us resolve the issue quickly.
- Yes, the BIKE system offers collaboration tools that allow you to work with other users. Visit the Collaboration section for more details.
If you need further assistance, please reach out to our support team through the Contact Support page. We are here to help you with any issues or questions you may have. Our typical response time is 24-48 hours.
This project is licensed under the MIT License. See the LICENSE file for more details.
Thank you for using ClinicalNLP Bike. We hope you have a great experience!