The Text Analysis and Machine Learning Group
Executive summary | What is Knowledge Management? | The team | Brief history | Results and accomplishments | Current graduate students | International, national and industrial cooperation
This document presents a five-year strategic plan of research in Knowledge Management at the University of Ottawa. We base our proposal on the track record of the existing Text Analysis and Machine Learning research group, active within the Department of Computer Science (and, subsequently, SITE) since 1988.
We demonstrate that, in terms of the standard measures of research quality, TAMALE is a world-class center of excellence in Data Mining, Text Mining, and Language Engineering. We observe that in recent years TAMALE has lost several of its most valuable researchers to other universities. These losses must be seen in the context of vigorous support of research in Knowledge Management elsewhere in Canada. We define and justify the resources necessary to sustain University of Ottawa’s position as a center of excellence in Knowledge Management.
Knowledge management is considered here as a research field that combines Data Mining, Text Mining and Language Engineering, and builds on the technologies of Databases, Data Warehousing and Knowledge Bases. Specifically,
- Data Mining is a field of research that extracts implicit, actionable knowledge from large, often unstructured data repositories. Data Warehousing addresses the organization of this data. For instance, data mining methods analyze gene expression data collected on DNA chips and builds a temporal model of that interaction.
- Text Mining, also known as Information Extraction from Text, identifies in large repositories of texts (e.g. news wire, Internet, virtual libraries) specific information of interest to the user. For instance, a text mining system will provide geneticians with knowledge of all interactions of two specific proteins ever mentioned in the bibliographic database MedLine within the last three years.
- Language Engineering combines techniques, representations, and algorithms that are collectively capable of advanced information processing involving natural language, e.g. answering questions about contents of a book, summarizing and compressing texts etc. For instance, a system reading texts on proteomics and summarizing them for a given research group, and ranking them with respect to their relevance for the group’s interests, is an example of a Language Engineering application.
Full-time faculty:
- Dr. Caroline Barriere (leaving); Language Engineering
- Dr. Natalie Japkowicz; Data Mining/Machine Learning, Natural Language Processing
- Dr. Iluju Kiringa; Knowledge Management
- Dr. Mario Marchand (leaving); Data Mining, Neural Networks
- Dr. Stan Matwin, Group Leader; Integrated Data Mining and applications
- Dr. Liam Peyton; Data Mining, E-commerce applications
- Dr. Stan Szpakowicz; Natural Language Processing, Language Engineering
- Dr. Marcel Turcotte; Data Mining, Bioinformatics
- Dr. Herna Viktor; Data Mining, Data Warehousing
Research Associate:
- Dr. Jelber Sayyad Shirabad; Machine Learning, Data Mining and their application to Software Engineering
Below we refer to the group members by their initials.
Research in machine learning started at University of Ottawa in 1988 when S. Matwin created the Ottawa Machine Learning Group. Initially it focused on analytical machine learning, as well as fundamental issues of inductive learning. R. C. Holte joined in 1989. With time, strong links were built with the Natural Language Lab led by S. Szpakowicz and in 1995 a joint group under the name of Text Analysis and Machine Learning (TAMALE) was launched. At different times, M. Marchand, K. Barker, C. Barriere, N. Japkowicz, L. Peyton, H. Viktor and I. Kiringa joined the group.
TAMALE has quickly gained international status in a number of areas of data mining and language engineering, e.g. Inductive Logic Programming, Performance Evaluation, Learning and NL, Syntactic Processing of NL. In 2000, L. Duquette from the Second Language Institute, with active participation of C. Barrière, S. Matwin, and S. Szpakowicz has created the Groupe de recherche en ingénierie de la langue (GRIL) which is a vehicle for collaboration of researchers from computer science, linguistics, and psychology, working on language engineering. The group has been awarded two major grants from SSHRC.
Current application areas in TAMALE span bioinformatics, computational linguistics, e-commerce and health.