Skip to content
TTGThesis Title Generator

TTG BS Computer Science (BSCS)

Computer Science thesis titles

A Computer Science thesis usually compares, adapts or evaluates an algorithm or model on a clearly defined dataset. Good titles name the technique, the task and the data, and the scope stays small enough to finish on a student laptop. Local data, like Filipino text, Philippine traffic or school records, makes a CS thesis feel grounded rather than generic.

Describe your thesis

Degree level

Topics, problems you've noticed, or people you want to study.

More options

Who you would study, e.g. Grade 10 students, MSME owners

Where, e.g. a public hospital in Davao City

e.g. no budget, one semester, online survey only

TTG runs on Google's free Gemini tier, so Google may use what you type to improve its products. Don't enter names, contact details or other personal information.

Your titles will appear here

Each suggestion comes with a short rationale, draft research questions, a suggested method and an honest note on how feasible it is, so you have something concrete to discuss with your adviser.

Not sure what makes a title strong? Read the guide to thesis titles.

Example titles for CS students

Six titles in the shape TTG suggests. Use them to see what “specific and feasible” looks like, not as titles to submit: your adviser will want a topic that's yours.

Comparative Evaluation of Naive Bayes and Support Vector Machine Classifiers for Detecting Scam Text Messages in Filipino and English

Scam texts mix Filipino, English and Taglish, which can confuse classifiers trained only on English. Comparing two well-understood models on a local dataset is a contained, reproducible thesis.

Research type
Experimental
Suggested method
Experimental comparison with k-fold cross-validation on a labeled SMS dataset, ≈3,000 messages
Feasibility
High feasibility. Runs on a laptop; collecting and labeling enough real scam messages is the slow part.
Ethics
Remove phone numbers and personal details from collected messages before labeling.

Sentiment Analysis of Taglish Product Reviews Using a Fine-Tuned Multilingual Transformer Model

Online shoppers in the Philippines write reviews in Taglish, which general sentiment tools handle poorly. Fine-tuning an existing multilingual model on local reviews tests a realistic improvement.

Research type
Experimental
Suggested method
Experimental evaluation against a baseline on a labeled review corpus, ≈5,000 reviews
Feasibility
Medium feasibility. Needs free cloud GPU time and careful annotation guidelines.

Development of an Offline Filipino Sign Language Alphabet Recognition App Using a Lightweight Convolutional Neural Network

Hearing people rarely learn Filipino Sign Language, and most recognition tools need a constant internet connection. An offline mobile model is a demanding but well-scoped build.

Research type
System development
Suggested method
Model training and on-device testing, with a usability evaluation among Deaf and hearing users, n≈30
Feasibility
Medium feasibility. Collaborating with the Deaf community for data and testing takes early planning.
Ethics
Recording participants' hands and faces requires informed consent and an accessible consent process.

Genetic Algorithm and Constraint Programming Approaches to Generating Class Schedules for a Small Private College

Class scheduling is still done by hand in many small colleges and takes weeks each term. Comparing two solvers on real constraints shows which is practical for a small school.

Research type
Experimental
Suggested method
Experimental comparison on one college's actual term data across problem sizes
Feasibility
High feasibility. Needs one term of anonymized scheduling data from a registrar.

Evaluation of an AI Chatbot for Answering Enrollment Questions at a State University Based on Answer Accuracy and Student Satisfaction

Enrollment offices answer the same questions thousands of times each term. Evaluating accuracy and satisfaction together shows whether a chatbot actually helps or only moves the frustration elsewhere.

Research type
Mixed methods
Suggested method
Accuracy test on 200 benchmark questions plus satisfaction survey (n≈120) and interviews
Feasibility
Medium feasibility. Requires the registrar's official policies to build an answer key.

Estimating Jeepney Route Congestion From Crowdsourced GPS Traces: A Case Study of Selected Routes in Cebu City

Route-level congestion data is scarce for public utility jeepneys, even though commuters feel it daily. Phone GPS traces from volunteers can map where delays actually happen.

Research type
Case study
Suggested method
Case study using GPS logging by volunteer commuters over four weeks, ≈20 volunteers
Feasibility
Medium feasibility. Volunteer retention over several weeks is the main risk.
Ethics
Location traces are personal data: anonymize start and end points and get informed consent.