A Revolutionary Approach to Humanities Computing?: Tools Development and the D2K Data-Mining Framework
J. Stephen Downie
jdownie@uiuc.edu
University of Illinois at Urbana-Champaign
John Unsworth
unsworth@uiuc.edu
University of Illinois at Urbana-Champaign
Bei Yu
beiyu@uiuc.edu
University of Illinois at Urbana-Champaign
David Tcheng
dtcheng@ncsa.uiuc.edu
University of Illinois at Urbana-Champaign
Geoffrey Rockwell
georock@mcmaster.ca
McMaster University
Stephen J. Ramsay
sramsay@uga.edu
University of Georgia
Introduction
A new set of humanities computing (HC) research projects are underway
that could revolutionize how the HC community works together to build,
use, and share HC tools. The set of projects under consideration all
play a role in the development work currently being done to extend the
D2K (Data-to-Knowledge)
[Note 1: See .]
data-mining framework into the realm of
HC. John Unsworth and Stephen J. Ramsay were recently
awarded a significant Andrew W. Mellon Foundation grant
[Note 2: See .]
to develop
a suite of HC data-mining tools using D2K and its child framework, T2K
(Text-to-Knowledge). Drs. Unsworth and Ramsay, along with
research assistant, Bei Yu,
are working closely with Geoffrey
Rockwell. Dr. Rockwell is the project leader for the CFI (Canada
Foundation for Innovation) funded project, TAPoR (Text Analysis Portal
for Research)
[Note 3: ]
, which is developing a text tool portal for
researchers
who work with electronic texts. J.
Stephen Downie and David Tcheng,
through their work in creating the International Music Information
Retrieval Systems Evaluation Laboratory
(IMIRSEL)
[Note 4: See .]
, are leading an international researchers group to
develop another D2K child system called M2K
(Music-to-Knowledge). This panel session demonstrates how all of these
projects come together to form a comprehensive whole. The session has
four major themes designed, through presentations and demonstrations,
to highlight individual the project components being developed and
their collective impact on the future of HC research. These themes are:
1.
D2K as the overarching framework
2.
T2K and its ties
to traditional text-based HC techniques
3.
M2K and its ties to multi-media-based HC
techniques
4.
The issues surrounding the HC community's development,
validation, distribution, and re-use of D2K/T2K/M2K modules.
Participants
J. Stephen Downie, Graduate
School of Library and Information Science (GSLIS), University of
Illinois at Urbana-Champaign (UIUC)
John Unsworth, GSLIS, UIUC
Bei Yu, GSLIS, UIUC
David Tcheng, National
Center for Supercomputng Applications (NCSA), UIUC
Geoffrey Rockwell, School
of the Arts, McMaster University
Stephen J. Ramsay,
Department of English, University of Georgia
Presentations, Demonstrations, and Discussions (in order)
Overview of the NORA (No One
Remembers Acronyms) project
John Unsworth
For decades, humanities computing
researchers have been developing software tools and statistical
techniques for text analysis, but those same researchers have not
succeeded in producing tools of interest to the majority of humanities
researchers, nor (with the exception of some very recent work in the
Canadian TAPoR project) have they produced tools that work over the
web. Meanwhile, large collections of web-accessible structured texts in
the humanities have been created and collected by libraries over the
last fifteen years. During that same time period, with
improvements database and other information technologies, data-mining
has become a practical tool, albeit one mostly used in business
applications. We believe data-mining (or more specifically,
text-mining) techniques can be applied to digital library collections
to discover unanticipated patterns, for further exploration either
through traditional criticism or through web-based text analysis.
Existing humanities e-text collections from Virginia, Michigan,
Indiana, North Carolina, and other research universities form the
corpus for the project. NORA brings NCSA's D2K data-mining architecture
to bear on the challenges of text-mining in digital libraries, with
special emphasis on leveraging markup, and on visualizations as
interface and as part of an iterative process of exploration.
Introduction to the D2K framework
David Tcheng
Released in 1999, D2K was developed by the Automated Learning Group
(ALG) at NCSA. D2K has been used to solve many problems for both
industry (e.g., Sears, Caterpillar, etc.) and government agencies
(e.g., NSF, NASA, NIH, etc.). Academic uses include bioinformatics,
seismology, hydrology, and astronomy. D2K uses a data flow
paradigm where a program is a network (directed graph) of processing
modules. Modules can be primitive, defined as a single piece of
source code that implements a single well defined task, or can be
nested meaning it is defined as a network of previously defined D2K
modules. Decomposition of programs into modules that implement a
well defined input-output relationship promotes the creation of
reusable code. Nesting modules into higher-level modules helps to
manage complexity. D2K parallelizes across any number different
computers by simply running a copy of "D2K Server" on each
available machine. The D2K software distribution comes as a basic D2K
package, with core modules capable of doing general purpose
data-mining, as well as such task-specific add-on packages as text
analysis (T2K), image analysis (I2K), and now music analysis (M2K).
Introduction to T2K
Bei Yu
Similar
to many data-mining tools, T2K has implemented a number of automatic
classification and clustering algorithms. Compared to the commercial
text mining tools, for example SAS Text Miner, T2K has richer NLP
preprocessing tools, especially after its integration with GATE. Tools
include: stemmer, tokenizer, PoS-tagger, data cleaning and named-entity
extraction tools. The clustering visualization is tailored for thematic
analysis. On one hand, T2K provides a text mining platform for the HC
community. On the other hand, T2K is also a platform to automate the HC
research results and thus facilitate their applications to the text
mining community in general. For example, most of the text mining tasks
are still topicality oriented, but the affect analysis has emerged in
the last couple of years. The affect of a document includes the
subjectivity/objectivity, the positive/neutral/negative attitude, and
the strength of emotions, etc. Some researchers have adapted stylistic
analysis techniques from HC to analyze customer reviews. The found
non-thematic features can also be used as predictors for document
genre, readability, clarity and many other document properties.
The TAPoR portal and D2K
Geoffrey Rockwell
TAPoR
has released an alpha of the portal and will have the beta ready by
June 2005. The portal is designed to allow researchers to run tools
(which can be local or remote web services) on texts (which can be
local or remote.) The TAPoR portal has been designed to work with other
systems like D2K in three ways:
1.
Particular tools or chains of tools can be published so that
they are available as post-process tool right in the interface of
another system. Thus one can have a button that appears on the
appropriate results screens of a D2K process that allows the user to
pass results to TAPoR tools.
2.
The portal has been released as open source and we are working on
models for projects to run customized versions of the portal that work
within their environment.
3.
The portal can initiate queries to remote systems and then pass
results to other TAPoR tools. Thus users can see tools like D2K (where
they have permission) within their portal account.
The Tamarind project and D2K
Stephen J. Ramsay
Tamarind began with the observation that the most basic text analysis
procedure of all — search — does not typically operate on the text
archive itself. It operates, rather, on a specially designed data
structure (typically an inverted file or pat trie index) that contains
string locations and byte offsets. Tamarind's primary goal is
to facilitate access to analytical data gleaned from large-scale full text
archives. Our working prototype of Tamarind, for example, can quickly
generate a relational database of graph properties in a text which can
in turn be mined for structural information about the texts in
question. Tamarind creates a generalized database schema for
holding text properties and allows you to specify this structure as one
that should be isolated and loaded into the database. Work is
proceeding on a module that will allow the user to load a Tamarind
database with millions of word frequency data points drawn from several
gigabytes of encoded data. Unlike existing tools, this newest
module includes information about where those counts occur within the
tag structure of the document (something that is impossible to do
without the raw XML). For the purposes of this project, we intend to
use D2K and T2K as the primary clients for Tamarind data stores.
The M2K project
J. Stephen Downie
M2K is being
developed to provide the Music Information Retrieval (MIR) community
with a mechanism to access a secure store of copyright-sensitive music
materials in symbolic, audio and graphic formats. M2K is a set of
open-source, music-specific, D2K modules jointly developed by members
of the IMIRSEL project and the wider MIR community. M2K modules include
such classic signal processing functions as Fast Fourier Transforms,
Spectral Flux, etc. In combination with D2K's built-in classification
functions (e.g., Bayesian Networks, Decision Trees, etc.), the M2K
modules allow MIR researchers to quickly construct and evaluate
prototype MIR systems that perform such sophisticated tasks as genre
recoginition, artist identification, audio transcription, score
analysis, and similarity clustering.
John Unsworth and J. Stephen Downie
will lead a wrap-up and future work open-forum discussion: For
ambitious, multi-institutional
projects like those presented in this panel many issues arise that can
affect the sustainability and impact of the projects. In particular,
the issues surrounding the HC community's development, validation,
distribution, and re-use of D2K/T2K/M2K modules will be
addressed.