Before joining Microsoft in 1997, Dumais was a researcher at Bellcore (now Telcordia Technologies), where she and her colleagues conducted research into what is now called the vocabulary problem in information retrieval. Their study demonstrated, through a variety of experiments, that different people use different vocabulary to describe the same thing, and that even choosing the "best" term to describe something is not enough for others to find it. One implication of this work is that because the author of a document may use different vocabulary than someone searching for the document, traditional information retrieval methods will have limited success.
Attributes | Values |
---|
rdfs:label
| |
rdfs:comment
| - Before joining Microsoft in 1997, Dumais was a researcher at Bellcore (now Telcordia Technologies), where she and her colleagues conducted research into what is now called the vocabulary problem in information retrieval. Their study demonstrated, through a variety of experiments, that different people use different vocabulary to describe the same thing, and that even choosing the "best" term to describe something is not enough for others to find it. One implication of this work is that because the author of a document may use different vocabulary than someone searching for the document, traditional information retrieval methods will have limited success.
|
dcterms:subject
| |
dbkwik:microsoft/p...iPageUsesTemplate
| |
abstract
| - Before joining Microsoft in 1997, Dumais was a researcher at Bellcore (now Telcordia Technologies), where she and her colleagues conducted research into what is now called the vocabulary problem in information retrieval. Their study demonstrated, through a variety of experiments, that different people use different vocabulary to describe the same thing, and that even choosing the "best" term to describe something is not enough for others to find it. One implication of this work is that because the author of a document may use different vocabulary than someone searching for the document, traditional information retrieval methods will have limited success. Dumais and the other Bellcore researchers then began investigating ways to build search systems that avoided the vocabulary problem. The result was their invention of Latent Semantic Indexing. In 2006, Dumais was inducted as a Fellow of the Association for Computing Machinery. In 2009, she received the Gerard Salton Award, an information retrieval lifetime achievement award.
|