DSCI 551 - Project Guideline
Spring 2025
ChatDB - Taking to Database Management Systems using Natural Language
In this project, you are asked to develop a natural language interface to one or more SQL/NoSQL database systems (depending on the number of people in your group). The interface should support the following functions for RDBMS (e.g., MySQL):
● Explore schemas and data of database: It should allow users to ask questions (in natural
language) to find out what tables there are in the database, what attributes a table has, and get sample data from a table.
● Query: it should accept natural language queries from users, and convert them into database queries (SQL), execute the queries in the DBMS, and return query results to the users. It should support the common constructs in SQL, including select, from, where, group by, having, order by, limit, and offset. It should also allow queries that involve joining multiple tables.
● Data modification: It should support insert, delete, and update requests in natural language. For example, add a new employee named John in the HR department; or update John’s age to 26.
Similarly, the interface should support the following functions for NoSQL databases (e.g., MongoDB):
● Explore schemas and data of database: It should allow users to ask questions to find out what collections there are in the database and get sample data from a table.
● Query: it should accept natural language queries from users, and convert them into database queries, execute the queries in the DBMS, and return query results to the users. It should support the functions in MongoDB: find (with projection), aggregate (with $match, $group, $sort, $limit, $skip, $project). Note that $match can be before and after $group. It should also allow queries that involve joining of two collections (using $lookup).
● Data modification: It should support insert, delete, and update requests in natural language. For example, add a new employee named John in the HR department; or update John’s age to 26.
ChatDB should convert the requests into MongoDB functions such as insertOne, updateOne, deleteOne, insertMany, etc.
Additional Requirements and Notes:
● You can form. a group of up to 3 people for your project.
o One-person group: You can develop a natural language interface (NLI) to either a RDBMS (except for sqlite) or a NoSQL database.
o Two-person group: You should develop NLI to a RDBMS and a NoSQL database.
o Three-person group: You should develop NLI to a RDBMS and a NoSQL
database. In addition, you should develop a Web browser-based UI for users to interact with ChatDB.
● You can use OpenAI or a large language model API for the projects. See the appendix for possible options and some notes which you might find helpful in your project implementation. You are encouraged to share your experiences and further resources on LLM API on Piazza!
● Use at least three databases to showcase the working of your ChatDB implementation. Note that you will need to have at least two tables/collections to demonstrate join.
Project deliverables:
● Proposal (due 2/7, Friday, 10 points): Detail your project plan, including type of database
system, databases, and how you plan to implement the functions stated in the beginning of this handout. Also list group members and their roles. Note that your project is a collaborative effort, and all members should contribute to the project. We will grade your project as a team effort and every member receives the same grade.
● Midterm progress report (due 3/7, Friday, 10 points): tell us your progress so far and the challenges you might have encountered. Note that any reformation of groups should be made before midterm progress and reported in the midterm progress report (also notify TA). It will be the responsibility of all group members to make sure the project will be a team effort after that.
● Demo (in-class, 4/21 and 4/23, 10 points): Give a live demo of your project. All project
members should be present during the demo, presenting his/her contribution. If you are absent, we will assume that you have not contributed to the implementation of the project.
● Final report (due on 5/9, Friday, 10 points): the final report should be comprehensive, details your design and implementation, and your learning experiences.
● Implementation (due on your demo time, 60 points): note your project should be fully
implemented before the demo. You should include in your final report a link to Google drive where you will upload your project codebase and documentations. Make sure you give access to your project folder.