Loading...
Please wait, while we are loading the content...
Similar Documents
Interactive Query Refinement for Boolean Search
| Content Provider | Semantic Scholar |
|---|---|
| Author | Harris, Greg Panangadan, Anand Prasanna, Viktor K. Ming-Hsieh |
| Copyright Year | 2014 |
| Abstract | Boolean search is still the method of choice for many kinds of professional search, such as constructing systematic reviews in legal and medical fields. It is effective for fast, high-recall document classification. Its drawback is the difficulty in crafting a Boolean query that captures semantically relevant documents. Ambiguous search terms lead to the inclusion of non-relevant documents. We propose a new exploratory document classification system that guides the user in creating a Boolean search query with both high recall and high precision. To increase recall, the system suggests semantically similar search terms. However, unlike other systems that use interactive query expansion, ours also includes a mechanism for removing non-relevant documents that inevitably get matched by an expanded query. The system explicitly identifies semantically ambiguous query terms that reduce the precision of the search. Removing non-relevant documents increases precision without reducing recall. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://ceng.usc.edu/~anandvp/pubs/boolean-search.pdf |
| Alternate Webpage(s) | http://ecs.fullerton.edu/~anandvp/pubs/boolean-search.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
