190 likes | 289 Views
Brett Hlavinka and Chris Aikens. Computer Science Search. Imagine…. You’re a CSCE Junior about to start upper-level courses You’re frustrated with howdy and its uselessness You don’t want to pay simply to look up professor reviews You want to find courses you have an interest in.
E N D
Brett Hlavinka and Chris Aikens Computer Science Search
Imagine… • You’re a CSCE Junior about to start upper-level courses • You’re frustrated with howdy and its uselessness • You don’t want to pay simply to look up professor reviews • You want to find courses you have an interest in
Sources of Information • Course information is spread across many sources • howdy • Course schedules • myEdu (formerly PickaProf) • Grade distributions and reviews • CS Department website • Course descriptions
Search Problems • Searching through courses is limited • howdy • Limit results with advanced search filters • myEdu (formerly PickaProf) • Course number or professor’s name • CS Department website • No searching
Motivation • Problem: Cannot easily determine which courses you want to take • Solution: A consolidated, searchable website for CS courses • View course schedules • Search over descriptions and reviews • Be free of charge • Enter Computer Science Search (CSS)
Related Work • MyEdu • Grade distributions and professor ratings/reviews • Starting at $10 a year
Related Work • Stanford’s Course Rank • Search, Review, Schedule, and Plan • Can be adopted by other universities
Related Work • AgProfessors.com • Course/professor grade distributions and reviews • Texas A&M specific (and free of charge)
Related Work • Summary • All related works incorporated good ideas • Searching • Reviews • Course Descriptions • Try to integrate related works and our own ideas • Idea: make a course search site specific to Texas A&M that helps students build their schedule • Should be easy to use and allow students to retrieve information
Methods Used • Scope: make the site CSCE only with static content to start • Collected course info • MyEdu reviews as base (plus some original) • Store various fields relevant to courses in an XML file • Use Digester to parse the XML file • Use Lucene to allow the user to search over courses and reviews
Demo • Search for Course Number 470 • Search for Name Hurley • Search for Time MWF • Search for All “Database” • Search for All “Aggies Roolz” • Browse by Number • Browse by Prof
Scenario 1 • Information need: retrieve relevant information for taking CSCE – 410 • AgProfessors: Search finds course and gives grade distributions • MyEdu: Search finds course and tells grade distributions and has professor reviews • CSS: Search finds course and gives general information about the course
Scenario 2 • Information need: retrieve relevant courses to the search “scheduling and memory allocation” • Try to find course that covers this material • Hope to find CSCE – 410 • AgProfessors and MyEdu fall short here • This is the reason behind CSS • The search returns CSCE – 410
Evaluation • All the searchable sites have useful information, the user needs to decide their information need • Grade Distributions • Professor Reviews • Course Scheduling/Reviews • CSS was made to find courses from a general search • We did not expect the other sites to perform well in this area
Analysis • Student values a site that is essentially all in one package • Stanford’s Course Rank has all the features that any student would need when constructing a course schedule • Search capability • Reviews • Built in scheduling • Planning
Conclusions • There is an obvious absence of searchable course listings • CSS is a site designed for finding courses you want • CSS could be further developed to implement more features • Writing reviews • Grade distributions • Course recommendations
Conclusions • Information should be consolidated • CSS is a consolidated site, combining course listings with searchable descriptions and reviews • Reduces time retrieving information, as well as the hassle of doing so