[Talk Summary 13] Concept Map Extraction from Textbooks

Shuting Wang, 5th year PhD student, from Computer Science department, Penn State University had the talk "Concept Map Extraction from Textbooks" on Dec 05, 2016. In the talk, she presented three parts of her work: extracting concept hierarchy from textbook, using prerequisite to extract concept maps from textbook, and using the concept maps for automatic assessment.

First of all, Shuting talked about how to extract concept hierarchy from textbook. In her work, she extracted important concepts in each book chapter using Wikipedia as a resource and then construct a concept hierarchy for that book. She presented the process to construct a concept hierarchy as follows:
  1. Build a concept dictionary from Wikipedia entities related to the topic of the book.
  2. Select concept candidates in the concept dictionary based on title and content similarity between a section (subchapter) in the book and Wikipedia articles.
  3. Construct the concept hierarchy from the table of content order of the book with considering two features local relatedness and global coherence.
 For the second part, she claimed that although being able to construct a concept hierarchy for textbook, the relationships between concepts are still simple, just considering the relations within a chapter. Hence, she introduced a framework for constructing a specific type of knowledge graph, a concept map from textbooks. The relations in this concept map are prerequisite relations among the concepts, which are derived by using Wikipedia. 

She noticed that different with previous work, which considered two problems key concept extraction and concept relationship identification separately. She proposed a framework that jointly optimizes these two problems and investigates methods that identify concept relationships. Shuting Wang's proposed concept map extractor receives the input including a digital book with a list of titles, chapter number and contents of all its chapter (e.i., each chapter contains one or more key concepts). The output is a concept map, which is presented as a set of triples in the form of concept A, concept B, and the relation between A and B (i.e., has a prerequisite relation or not).

In the end of the talk, Shuting briefly presented an application of concept maps for automatic assessment. Her group proposed the idea of utilizing prerequisite concept map to detect learning gaps for efficient learning. In addition, they also derived a top-k concept selection algorithm which allows students to view different numbers of concepts, and help to select the most suitable concepts to assess learner's knowledge in case the learners do not which concepts they should start for assessment.

Room 828, School of Information Sciences
University of Pittsburgh 


Popular posts from this blog

FolkTrails: Interpreting Navigation Behavior in a Social Tagging System

[Talk Summary] Machine Learning and Privacy: Friends or Foes?

Collaborative Filtering in Social Tagging Systems Based on Joint Item-Tag Recommendations