Building Multilingual Semantic Networks with Non-Expert Contributions over the Web Page: 1
The following text was automatically extracted from the image on this page using optical character recognition software:
Building Multilingual Semantic Networks with Non-Expert
Contributions over the Web
Southern Methodist University
University of North Texas
University of Ottawa
Ottawa, ON, Canada
We present a system that allows non-expert Web users
to contribute towards building a multilingual lexical re-
source. Our study focuses on the Romanian-English
language pair, and the target resource is a Romanian
WordNet strongly connected to the English WordNet.
We use a bilingual dictionary, a monolingual definition
dictionary and documents on the Web to build synsets,
attach them a gloss, and provide some examples. The
results of our semi-automatic acquisition system are
judged by two human judges, and they are compared
to automatic approaches to building a Romanian Word-
In order to obtain a system that provides expertise in
a specific domain, the knowledge of that domain must
be made available in a format that the system can use.
Developers of software often do not have the knowledge
of such specific domains, and experts in the field do not
have the knowledge to create such a knowledge base.
This has led to a new trend, in which software devel-
opers write tools that allow experts to readily formalize
their knowledge through the system provided, which
then encodes this input in a format that a system can
Language is a field that all people are experts in. We
offer them RSDNET - a tool freely available on the In-
ternet, with a friendly interface, through which Web
contributors participate in the construction of a multi-
lingual semantic network, by validating automatically
suggested synonym sets.
We present in this paper the paradigm behind this sys-
tem, the implementation and the interface, the role of
the user, and an analysis of the results obtained so
far. The results gathered were analyzed by two hu-
man judges, and compared to results obtained in other
2. RELATED PROJECTS
The idea of harnessing the knowledge of experts in a
particular field in order to gather data has found many
The Rapid Knowledge Formation project  is geared
towards providing experts in various fields with tools
that allow them to encode their knowledge in an in-
tuitive way, without needing to acquire programming
skills. This is realized by using a graphical interface,
which the experts manipulate to form and link concepts
Collecting data over the Web for a variety of AI ap-
plications is a relatively new approach. The basic idea
behind the broad Open Mind initiative  is to use the
information and knowledge obtainable from millions of
Web users to create more intelligent applications. Open
Mind projects include our own effort - Open Mind Word
Expert  - to build lexically annotated corpora through
volunteer contributions. They also include Open Mind
1001 Questions , which acquires knowledge and Open
Mind Common Sense , a system that collects com-
mon sense statements from Web users.
The domain of expertise our project is focused on is
language. All native speakers of a language are ex-
pert users of their mother tongue. A structured system
can help them focus on particular aspects, and harness
their knowledge towards the construction of interesting
resources. OpenMind Word Expert provides such a sys-
tem, allowing people all over the world to contribute
towards building a corpus annotated with semantic in-
WordNet  is a lexical resource that is used frequently
in the NLP community for word-sense disambiguation,
question answering and summarization, and other tasks.
Here’s what’s next.
This paper can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Paper.
Ayewah, Nathanial; Mihalcea, Rada, 1974- & Nastase, Vivi. Building Multilingual Semantic Networks with Non-Expert Contributions over the Web, paper, November 2003; (digital.library.unt.edu/ark:/67531/metadc30947/m1/1/: accessed October 21, 2017), University of North Texas Libraries, Digital Library, digital.library.unt.edu; crediting UNT College of Engineering.