Data processing: database and file management or data structures – Database design – Data structure types
Reexamination Certificate
2000-03-31
2002-12-03
Nguyen, Cam Linh T. (Department: 2171)
Data processing: database and file management or data structures
Database design
Data structure types
Reexamination Certificate
active
06490577
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a search engine that captures and stores user activity data and uses the stored data to refine search rankings and to methods implemented on a computer for capturing, storing and analyzing user activity data and refining rankings in subsequent searches.
More particularly, the present invention relates to a search engine that after displaying search results captures and stores user activity data. The user activity data is then analyzed and used to refine ranking in search results in subsequent searches where the ranking protocol will include user activity data, record based data and relationships derived therefrom. The accumulated user activity data can be used to improve search results, improve search efficiency and improve list rankings so better service the user.
2. Description of the Related Art
Search engines are well-established in the industry. The search engines provide users with a procedure for finding data in a database. Unfortunately, the current search engines use a ranking for displaying records that meet a particular query in whole or in part which is derived from the data in the record that meet the query criteria. Thus, if a record includes many references to a given search term, it may be placed higher on the list which the user must scroll through. There simply is no user derived data other than the search query to aid in ranking records. Thus, there is a need in the art for a search engine that incorporates into its ranking criteria the user activity data.
ABBREVIATIONS AND DEFINITIONS
The following words and/or abbreviations which may appear throughout this disclosure will have the meaning set forth below:
URL means universal record location.
UAD means user activity data.
RSD means record specific data.
RP means ranking protocol.
DB means database.
URLDB means universal record location database.
DPU means a digital processing unit.
CPU means a central processing unit.
Token means a character string (keyword), punctuation and operators.
Keyword means a unique character string which appears in a database entry.
Operator means any Boolean operator for connecting keywords.
Query means a collection of keywords, punctuation and operators that represents a request for data from a user.
SUMMARY OF THE INVENTION
Devices and Software Implemented in a DPU
The present invention provides a search engine component implemented in a DPU including a packet processing module that converts each packet of as search to at least one transaction and processes each transaction. The engine component also includes a database including records, keywords, user activity data, and a user activity data updating routine for updating user activity data. The engine component also includes a ranking protocol constructed from record specific data and user activity data for ranking retrieved records from the records database.
The present invention also provides a search engine component including a packet processing module that converts each packet of a search or query into at least one transaction and processes the transaction. The engine components also includes a database module that includes at least one records database that may include user activity data. The database module includes a keywords database including user activity data and a user activity data updating routine for updating user activity data. The engine component also includes a ranking protocol constructed from record specific data and user activity data for ranking retrieved records from the records database.
The present invention also provides a search engine component including a packet processing module that converts each packet of a search or query into at least one transaction and processes the transaction. The engine components also includes a database module that includes at least one records database, where the database module includes a keywords database; a user activity data database, and a user activity data updating routine for updating user activity data. The engine components also includes a ranking protocol constructed from record specific data and user activity data for ranking retrieved records from the records database.
The present invention provides a search engine component including a packet processing module that converts each packet to at least one transaction and a ranking protocol including record specific data and user activity data for ranking retrieved records from a records database. The search engine component also includes a database module including at least one records database, a keyword database, a user activity data database, and a user activity data database updating routine for updating the user activity data database. The search engine component also includes a transaction processing module for processing transactions. The transaction processing module includes a new query processing routine for retrieving records satisfying the query and ranking the retrieved records according to the ranking protocol; a selected record processing routine including a first user activity data activity value and a count incrementer; a record range processing routine including a second user activity data activity value and the count incrementer for identifying retrieved records viewed by a user; a terminal activity processing routine including a third user activity data activity value and the count incrementer.
The search engine components of this invention can further include a query refinement module for identifying clustering of records based on user activity data and RSD so that user queries can have improved relevance and context.
The search engine components can be combined with other search engine components such as a user interface, RSD processing routines, communication and I/O routines, and like to produce complete search engines of the present invention.
The present invention provides a database module including at least one database including records, keywords, user activity data, and a user activity data updating routine for updating user activity data.
The present invention includes a ranking protocol including RSD and user activity data for ranking retrieved records from the records database.
Methods and Processes
The present invention also provides a method implemented in a search engine of the present invention stored in a DPU for capturing UAD involving receiving a user response, processing the user response, computing response UAD according to an user activity protocol if the response is not a new query, and updating UAD in appropriate databases in the engine.
The present invention also provides a method implemented in a search engine of the present invention stored in a DPU for preprocessing a user query involving receiving a new query, determining query ambiguities, presenting to the user ambiguity resolving selections, and processing a user selected ambiguity resolving selection.
The present invention also provides a method implemented in a search engine of the present invention stored in a DPU for updating a ranking protocol involving updating UAD based terms in the protocol.
The present invention provides a method implemented in a search engine of the present invention stored in a DPU for searching involving receiving a user response and processing the user response. If the response is a new query, then records satisfying the query are assembled into a list. The assembled list is then ranked according to a ranking protocol comprising RSD and UAD terms and the ranked list is presented to the user. If the response is not a new query response or a termination response, then response UAD are computed according to a user activity protocol and corresponding UAD are updated in appropriate databases. The updated UAD is then used to refine the ranking protocol. The method is continued until the user session is completed. The method may also include the query preprocessing method described above.
REFERENCES:
patent: 5796952 (1998-08-01), Davis et al.
patent: 5864845 (1999-01-01), Voorhees et al.
patent
Nguyen Cam Linh T.
PolyVista, Inc.
Strozier Robert W.
LandOfFree
Search engine with user activity memory does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Search engine with user activity memory, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Search engine with user activity memory will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2994764