Automatic Syllabus Classification JCDL – Vancouver – 22 June 2007 Edward A. Fox (presenting co-author), Xiaoyan Yu, Manas Tungare, Weiguo Fan, Manuel Perez-Quinones, William Cameron, GuoFang Teng, and Lillian (“Boots”) Cassel
Why Study the Syllabus Genre? ► Educational resource ► Importance to the educational community Educators Students Self-learners ► Thanks to NSF DUE grant (personalization support for NSDL)
Where to look for a specific syllabus? ► ► Non-standard publishing mechanisms: Instructor’s website CMSs (courseware management systems, e.g., Sakai) Catalogs ► ► Limited access outside the university ► Search on the Web Many non-relevant links in search results
Syllabus Library ► Bootstrapping Identify true syllabi from search results Store in a repository Develop tools & applications ► Scaling up Encourage contributions from educational communities
An Essential Step towards Syllabus Library: Classification ► Classification Objects: Potential syllabi in Computer Science: search on the Web, using syllabus keywords, only in the educational domains ► Class Definition ► Feature Selection ► Model Selection ► Training and Testing
Four Classes Noise
Full Syllabus
Partial Syllabus
Entry Page
Noise
Syllabus Components ► ► course code ► ► title ► ► class time& location ► ► offering institution ► ► teaching staff ► ► course description ► ► objectives ► web site ► prerequisite ► textbook ► grading policy ► schedule ► assignment ► exam and resources
Features ► 84 Genre-specific Features the occurrences of keywords the positions of keywords, and the co-occurrences of keywords and links ► ► A series of keywords for each syllabus component
Classification Models ► Discriminative Models Support Vector Machines (SVM) SMO-L: SMO-L: Sequential Minimal Optimization, accelerating the training process of SVM SMO-P: SMO with a polynomial kernel ► Generative Models Naïve Bayes (NB) NB-K: Applying kernel methods to estimate the distribution of numeric attributes in NB modeling
Evaluation ► Training corpus: 1020 out of the potential syllabi ► All in HTML, PDF, PostScript, or Text ► Manual tagging on the training corpus Unanimous agreement by three co-authors ► Evaluation strategy: ten-fold cross validation ► Metrics: F 1 (an overall measure of classification performance)
Results w. random set Best items are in purple boxes. Acc tr : Classification accuracy on the training set.
Results (Cont’d) ► SVM outperforms NB regarding our syllabus classification on average. ► All classifiers fail in identifying the partial syllabus class. ► The kernel settings for NB are not helpful in the syllabus classification task. ► Classification accuracy on training data is not that good.
Future Work ► Feature selection Add general feature selection methods on text classification e.g., Document Frequency, Information Gain, and Mutual Information Hybrid: combine our genre-specific features with the general features
Future Work (Cont’d) ► Syllabus Library Welcome to Share your favorite course resources – not limited to the syllabus genre. ► Information Extraction Semantic search ► Personalization
Summary ► Towards a syllabus library Starting from search results on the web Classification of the search results for true syllabi ► SVM is a better choice for our syllabus classification task. ► Towards an educational on-line community around the syllabus library
Q & A