Data processing: database and file management or data structures – Database design – Data structure types
Reexamination Certificate
1998-10-14
2002-09-17
Corrielus, Jean M. (Department: 2172)
Data processing: database and file management or data structures
Database design
Data structure types
C707S793000, C707S793000, C345S215000
Reexamination Certificate
active
06453312
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates generally to an improved system and method for aiding users in the development of a query string; and more specifically, to a system and method for allowing a user to control the development of a concept-based natural language search query by controlling the manner in which a hierarchical concept tree is structured and traversed.
2. Description of the Prior Art
Today's data processing systems are capable of storing large volumes of data. The density of storage devices continue to increase at the same time the prices for such devices are falling. In addition, current networking capabilities allow multiple storage devices and file servers to be interconnected so that databases can be shared across systems. As a result, users of current computer systems are provided with access to an unprecedented amount of information.
The internet is a prime example of this information explosion. It is estimated that between 30 and 50 million pages are currently available to users of the internet. In addition to this publicly-available documentation, many users are also provided with other proprietary sources of information, such as those that are available via corporate intranet sites. Information may be obtained from still other sources, such as newsgroups.
To take advantage of the large amount of information made available by technological advances, the information must be readily accessible. Users must be able to locate and retrieve the documents that are needed in a timely manner. To do this, information retrieval systems must be developed that allow users to identify the best or most relevant information associated with a user request.
Many challenges exist when developing an information retrieval system that is capable of aiding users in finding meaningful information from a large body of electronic data. Often times, the users of such systems are only familiar with general topics of interest, and are not able to specify the actual terminology used within the textual information that is relevant for a desired topic. Additionally, the user may not be familiar with the keyword descriptors used to index the documents. As a result, the user-provided query may be incomplete or inaccurate. Other factors, for example regional language dialects, may further influence the construction of a search query. For all of these reasons, the information retrieved during a search may only yield a small number of documents that are actually relevant.
One way to improve search results is to provide a mechanism for automatically expanding a user-provided query string to include terms that do not appear in the query, but which may correspond to, or be associated with, the user-provided query terms. A system of this type is disclosed in U.S. Pat. No. 5,265,065 to Turtle. This patent describes a system in which words of a natural language input query are replaced with phrases from a database in a manner that expands the query. The expanded query is then utilized for information retrieval. The problem with many prior art systems of this type is that no opportunity is provided for the user to interactively participate in the query development process. If the search is unsuccessful, the user must simply re-execute the search with a different query string.
One method which does provide an iterative technique for allowing a user to interactively refine a query is provided in U.S. Pat. No. 5,278,980 to Pedersen et al. This patent describes a process whereby a user-provided query is developed into a search string that is further used to locate a list of matched phrases from a corpus of documents. Words from the returned phrases that are not included in the original query can be used to refine the query. This process can be repeated to retrieve documents that are increasingly focused on the desired topic. Although this system provides an opportunity for the user to exert control over the query development, the user intervention is only allowed after a search has been performed. This may waste processing resources, and does not provide any insight into the manner in which the user-provided query is developed into the actual search string.
Other types of search tools have been developed which seek to expand user-provided queries by employing context-based analysis techniques. These search tools analyze a phrase provided by a user in attempt to “understand” the user's intent. The search tools place various search terms within the context of other search terms so that the concept behind the query can be determined. While these types of tools can result in the retrieval of more relevant documents, these tools have the above-mentioned disadvantage of not allowing a user to participate in the search development process. That is, these tools utilize predetermined algorithms that can not be influenced by the user. Once the user provides an initial query string, additional query analysis and development is under the control of the tool, and the user is not provided any ability to control the algorithm or the lexicon employed in increasing the scope of the search. Thus, the number of irrelevant documents retrieved may actually be increased instead of decreased.
Another problem with existing search tools is that the user is not allowed to specify, with any degree of definiteness, the extent to which a query should be expanded. For example, a user may want to specify a given topic like “Mexican Cooking” for document retrieval. The user may further want the same search to return documents on growing peppers. Prior art search engines do not allow the user to exert control over the precise scope of the search so that documents concerning multiple related, yet distinct, topics can be retrieved using the same search.
Yet another drawback associated with prior art tools involves the limited visibility provided into the actual query development process. Because the user is not allowed to view the manner in which search expansion is accomplished, it is difficult to determine how a query string should be revised to retrieve more relevant documentation. If a search is unsuccessful, the user is left to guess as to how the query might be modified to obtain more meaningful results.
Still another problem with current document retrieval systems is that both the lexicon and the algorithm employed to expand a search is fixed. That is, the user is not allowed to modify the content or the arrangement of the index employed to develop a query. Thus, query development can not be customized to account for professional terminology, business or company-specific acronyms, newly-coined expressions, and the like, that may be included in a particular user's corpus of documentation.
What is needed is a flexible search system that allows users to closely control the expansion of a search query. The system should also provide the capability for a user to modify the manner in which particular query expansions. This capability should include the ability to modify both the system lexicon, and the lexicon organization. The system should allow users to add user-specific terminology such as regional jargon, slang terms, foreign language representations, or any other particularized phrases included in the body of material to be searched.
OBJECTS
It is a primary object of the current invention to provide an improved system for aiding users in developing search queries;
It is another object of the invention to provide a system that allows users to interactively control the manner in which a search query is expanded;
It is yet another object of the invention to provide a system wherein an iterative, interactive process is utilized to allow a user to expand a search query while exploring the lexicon of the system;
It is still another object of the invention to provide a system for cataloging search terms in a manner that is controllable by users;
It is yet another object of the invention to provide a system for aiding in the development of search queries, wherein the system has an ex
Goiffon David A.
Hartmann Gerald E.
Johnson David R.
Atlass Michael B.
Corrielus Jean M.
Fleurantin Jean Bolte
Johnson Charles A.
McMahon Beth L.
LandOfFree
System and method for developing a selectably-expandable... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method for developing a selectably-expandable..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for developing a selectably-expandable... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2824840