Also, keyword searches in these tools are prone to false drops. For instance, I was once given a demonstration by someone who was showing off his company's discovery tool, and he typed in the keyword search "East of Eden." It brought in results showing hits from books about Steinbeck and an encyclopedia article about the novel.
When your search is first displayed, it is surrounded by related terms, so there is much less chance of a false drop as with the East of Eden example.
In reference [18] reached 85.5% pass rate with 91
false drops, but we can find that the performance of our method is much better than Schneiderman's method; it can reach a high detection rate with less
false drop.
Controlled terms that are tailored to a corpus minimize
false drops. The ability to perform date limiting eliminates the results list that mixes old and new content indiscriminately.
Although this approach gives some space saving, the number of
false drops will definitely be increased due to sparse signature files.
Using PLSA, the company says MindServer Retrieval has the capability to "understand" the context of search terms utilizing machine learning technology, which results in far fewer incorrect results or
false drops.
The inclusion in the database of a good selection of items in full text greatly expands the access to Peel materials (with an attendant expansion of frustration due to "
false drops," but an even greater expansion of serendipitous finds).
In addition, the inverted file can only answer the general query "Find all images in which the query objects are present." Thus, the inverted file will suffer from a large number of
false drops if spatial constraints are included as part of the query.
Better retrieval can be achieved because of the semantic values attached to the terms; this would help reduce the
false drops. Document surrogates can be prepared automatically according to the prescribed rules of the concerned indexing system--e.g., according to the principle of generating the index entries in PRECIS (Austin & Dykstra, 1984).
When we look through the list of documents or sites returned from a search engine there are always a few errors, called "
false drops." When we can work out to our satisfaction how the
false drop occurred (perhaps a misspelled word), then we feel we can trust the search engine.
In xrefer you may also find
false drops that are unavoidable, especially in single-word searches, such as those results that list people whose first, middle, or last name is Hercules, not to mention the C-130 Hercules military airplane.
In contrast with results from general Web search engines and even Yahoo!, the leading Web directory service, searching on LookSmart seems to retrieve few or no
false drops. The company clearly has a commitment to "bull's-eye" searching.