Patent Application
20080195586
The present disclosure generally relates to data processing. More particularly, the present disclosure relates to performing searches using context information representative of a relationship between the user performing a search and a document being searched.
The amount of searchable data on the Internet is increasing on a daily basis. Using a search engine or tool to find the right information during a search has become increasingly difficult due to large amounts of searchable data now available electronically. As a consequence, users are becoming inundated with search results, many of which are not relevant to the user's search. One way to manage search results is to use direct, bidirectional relationships between a user and the results of the search. For example, search results may be limited by the location of the user and the location of the document being searched (e.g., limiting search results to documents having the same country as the country of the user performing the search). This approach, however, may not be effective since the actual location of the user and the actual location of the document may not be available or accurate. Alternatively, a user may avoid being inundated by a large quantity of search results by initiating a deep, structured search including many structured search terms. This approach, however, is burdensome to the user performing the search and does not provide any ranking mechanism for the search results. Providing mechanisms to enable meaningful search results continues to be a challenge for developers.
The subject matter disclosed herein provides methods and apparatus, including computer program products, for performing searches using context information representative of a relationship between the user performing a search and a document being searched. The context information may be used to compute the relevancy ranking of the documents identified by any search results.
In one aspect, there is provided a computer-implemented method. The method may include receiving a keyword to perform a search for one or more documents. The context information representative of a relationship between a user performing the search and the one or more documents being searched may be determined. The relationship may be representative of an organizational relationship between the user and the one or more documents. The results of the search may be provided based on the determined context information.
Variations may include one or more of the following features. The documents may be implemented as data stored in a file system (e.g., unstructured data, such as Word™ and PDF documents) and structured data stored in a relational database (e.g., business objects). The relationships between the user performing the search and the authors of the documents may be used to rank as well as limit the search. Moreover, the relationships may be defined by the organizational relationship between the user and the authors. The relationships may be determined based on context information representative the user and the authors that created the documents included in the one or more results. The context information may be determined as a distance metric representative of the organizational relationship between the user performing the search and the documents being searched. The distance metric may be determined from a first organization corresponding to the user performing the search and a second organization corresponding to at least one of the documents being searched. The distance metric may be determined by determining a number of hops from the first organization to the second organization. The results may be ranked based on the determined context information. The search may be limited based on the determined context information.
The subject matter described herein may be implemented to realize the advantage of providing to a user relevant search results by using context information associated with the user performing the search and the documents being searched. The context information may be used for ranking, e.g., sorting the result items (e.g., documents) to provide an indication of their relevance to the user. For example, the user's relationship to the authors of the resulting items (e.g., documents identified by the search) may be used to determine the relevance of the resulting items.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive. Further features and/or variations may be provided in addition to those set forth herein. For example, the implementations described herein may be directed to various combinations and subcombinations of the disclosed features and/or combinations and subcombinations of several further features disclosed below in the detailed description.
In the drawings,
Like labels are used to refer to same or similar items in the drawings.
System 100 includes a processor, such as a computer 110, coupled through communication link 150 (e.g., the Internet or an intranet) to another processor, such as server 190. The computer 110 may further include a user interface 120 for interacting with server 190 and a search application 142. The search application 142 may perform one or more of the following functions: determining context information corresponding to a user performing a search; determining context information corresponding to the documents being searched; receiving a keyword used to perform a search for one or more documents; searching indexes of documents to perform the search; providing search results based on the determined context information; ranking search results based on the determined context information; searching based on the determined context information and a keyword; determining the context information as a distance metric representative of the relationship between the user performing the search and the documents being searched; and determining the distance metric using human resources information, such as an employee directory or an organizational structure. In some implementations, search application 142 may be implemented as a search engine.
Although context information (including information representative of the organizational relationship between a user and the documents) may be obtained from human resources data including the position of the user performing the search and the position of others associated with documents being searched (e.g., an author of the document, a creator of a document, a modifier of a document, an owner of a document, a custodian of a document, and the like), the organizational relationship information may be obtained from sources other than human resource data and may take other forms. For example, the organizational relationship information may be obtained from an address book, an employee directory, keywords listed in an employee directory, and any other information that may be used to determine a relationship between the user (i.e., the searcher) and the document (e.g., the author, creator, owner, editor, etc).
At 220, the search application 142 determines context information for the user of user interface 120. For example, search application 142 may determine the user's position within an organization based on the user's login identifier. The search application 142 may also determine context information associated with documents being searched. For example, the search application 142 may determine the positions within an organization of one or more authors of the documents being searched. In this example, the context information is in the form of an organizational relationship between the user's position and the authors' position with the organization.
In some implementations, determining context information may also include search application 142 determining a so-called “distance metric” between the user performing the search and the document being searched.
In some implementations, keywords from an employee directory are used as context information to determine a relationship between a user performing the search and the author of the document(s) being searched. When that is the case, the correspondence of keywords in an employee directory for the user and author may be used. For example, if Nancy (
In some implementations, an address book is used as context information to determine a relationship between a user performing the search and the author of the document(s) being searched. When that is the case, several attributes may be used to determine a relationship between an author of the document and the user performing the search. For example, the relationship may be determined from geographical information including attributes like “office location” and/or “office number.” The relationship information may also be determined from organizational information in the address book including one or more of the following: the name or phone number of a secretary (or supervisor), an in-house mail address, a cost center, and an organizational identifier. Moreover, the number of corresponding attributes may be used as a distance metric. For example, the more attributes from the address box that match between the user and the author, the closer the relationship is in terms of distance. The closer the relationship, the higher the corresponding document is ranked when compared to other documents having fewer matching attributes. In addition, attributes may be weighted, so that some attributes are considered more than others are. For example, if a user and the author of a document have the same attribute “secretary” (e.g., both sharing the same secretary), that document would be ranked higher when compared to similar documents without a searching user and document author having the same secretary attribute. Alternatively, using weighted attributes, two documents sharing the same secretary between user and author may be ranked higher than documents from people sharing the same building by weighing the attribute “secretary” twice as much as the attribute “building”. In this example, the attribute secretary is more relevant in ranking and finding documents that merely working in the same office building.
Referring again to
In some implementations, the search application 142 may limit the search results by using context information as a search term to limit the search results. Specifically, the search application 142 may search using the keyword (provided by the user) and search for other context information. The context information may be in the form of metadata and may be categorized based on attributes. For example, a document may include metadata categorized as attributes, such as filename, file location, created by, modified by, owned by, authored by, category, keyword(s), and the like. The search may include the keyword as well as context information. Returning to the above example, the search terms may include “Budget 2007” and “authored by Paul.” The context information in this example limits the user's search of “Budget 2007” to documents where “Paul” is an author, so that the search results are more likely to be relevant to the user of user interface 120. The usage of metadata including context information may be done explicitly by the searching user (as described in the previous example), but usage of the metadata may also be done implicitly by search application 142. In the implicit case, the search application 142 may determine a list of people having the same attribute (e.g., manager, organizational identifier, such as department) and then add the list of people to the search query. For example, if Nancy is searching for documents including the keyword “Budget 2007,” the search application 142 may automatically search for “Budget 2007” AND “author=Paul OR author=Nancy OR author=Karen OR author=Jo” to limit the number of search results to documents from authors in “Department 1” (
Although the above description refers to searching “documents,” searching documents may also include searching indexes of documents, such as indexes created by search applications. Moreover, the documents may include any item including files, web pages, objects, images, audio, word processing documents, HTML pages, code, and structured data (e.g., business objects stored in a relational database). A business object (BO) represents structured data and/or methods. The term business object (BO) represents an object, such as a data structure including data and operations, of significance to a business. Examples of business objects include a purchase order, a sales order, a flight reservation, a shipping order, customer information, employee information, material master, business partner information, invoice information, business objects like business partner in an ERP system, and the like.
Referring again to
Communication link 150 may be any type of communications mechanism and may include, alone or in any suitable combination, the Internet, a telephony-based network, a local area network (LAN), a wide area network (WAN), a dedicated intranet, wireless LAN, an intranet, a wireless network, a bus, or any other communication mechanisms. Further, any suitable combination of wired and/or wireless components and systems may provide communication link 150. Moreover, communication link 150 may be embodied using bidirectional, unidirectional, or dedicated communication links. Communication link 150 may also implement standard transmission protocols, such as Transmission Control Protocol/Internet Protocol (TCP/IP), Hyper Text Transfer Protocol (HTTP), SOAP, RPC, or other protocols.
Server system 190 may include one or more processors, such as computers, to interface with other computers, such as computer 110, and/or programs, such as user interface 120. The search application 142 may be implemented as a program, group of programs, and/or component, i.e., a small binary object (e.g., an applet) or program that performs a specific function and is designed in such a way to easily operate with other components and programs.
In some implementations, the search application 142 may enable and disable the use of context information during a search. Moreover, although
At 430, the search application 142 provides the results (e.g., unspecified hits) of the search, such as any documents including the keyword being searched. The results may also include a description of each document, metadata associated with each document, and a location of each document.
At 440, the search application 142 may retrieve one or more attributes for the results of 430. For example, the documents may each include metadata. The metadata may include context information organized into attributes, such as an owner of a document, the last editor of a document, a creator of a document, customer information, supplier information, competitor information, department information, organizational information, a cost center, a project assignment, a job function, a job, a role, an employee type, a position in an organization, and/or self-assigned keywords.
At 450, the search application 142 may determine users associated with the search results. Table 1 below shows examples of search results (e.g., Document 1-Document 4) and users associated with the documents.
At 460, the search application 142 may then determine context information for each of the users associated with the documents. For example, the search application 142 may utilize context information, such as human resources information, address book information, employee directory, and organizational information, to determine context information for each of the users associated with the documents. In the example of Table 1, the search application 142 may determine the position in an organization of each user.
At 470, the search application 142 may then associated the context information with the attributes. Table 2 below depicts such an association.
At 480, the search application 142 may determine context information, such as human resources information, address book information, employee directory, and organizational information, for the user performing the search. For example, if the user performing the search is Nancy, her position in the organization chart of
At 490, search application 142 may then associate the context information of the user performing the search with attributes. In some implementations, the attributes for the user performing the search are the same type of attributes as the ones determined at 470. For example, if the attribute “created by” were used at 470, the same attribute would be used at 490. In this example, the attribute to be determined would be the department, corresponding to Tables 1 and 2. In this example, Tables 1 and 2 represents a so-called “positive/negative decision” on the department of the searching user and the author of the documents, as there is a relation between the departments preconfigured in the address book. In this case, the distance metric may be determined using the hop method defined above or as a binary decision (e.g., a distance of “1” if user and author are in the same department, otherwise a “0”). Alternatively, the manager (or department) of the searching user and of the document authors may be determined to enable a distance metric to be calculated by counting the number of hops along the organization chart.
At 495, the search application 142 ranks the results of the search by determining the relationship between the context information of the user performing the search and the context information associated with the documents. The search application 142 may determined a distance metric to determine the relationship between the context information of the user performing the search and the context information associated with the attributes of a document. For example, if Nancy is the person performing the search and the attribute is “department” and/or “created by,” a distance measurement from Nancy to the documents may be as follows: 0 (Document 1), 2 (Document 2), 4 (Document 3), and 6 (Document 4). The documents would, in this example, be ranked as follows: Document 1, Document 2, Document 3, and Document 4. In some implementations, the ranking at 495 may be performed by first computing the distance metric between a user and an author, and then ranking the results.
In some implementations, multiple attributes are used (e.g., created by and last edited by). When that is the case, the distance metric for each attribute is determined and then the distance metrics are combined. For example, a distance metric is determined for created by and another distance metric is determined for last edited by; the two distance metrics are then combined to provide an overall distance metric for ranking search results.
At 498, the search application 142 provides the ranked search results to user interface 120, so the user may view the results of the search.
At 510, the search application 142 may determine context information for a user performing a search by extracting the context information from, for example, human resources information, address book information, employee directory, and/or organizational information. To extract such information, the user performing the search may be required to first login to clearly authenticate and identify the user. The user's identity (e.g. the identity (ID) from a logon ticket) may be used to retrieve further information for that user, such as information from a human resources (HR) system (e.g., employee information database) or an employee directory. The HR system may include a data set for each employee. This data set may contain the following attributes: organizational information (e.g., unit, department, group, etc.), where the employee works in (e.g., office location), and the manager to which the employee reports, and a cost center to which the employee expenses are booked. For example, the HR system may be configured to include the function “getEmployeeData” for an employee. When the user performing the search is an employee, information associated with the aforementioned attributes may be retrieved using that function. In the example of Table 2 above, the search application 142 may use the retrieved information to determine the position of the user in an organization, such as the organization chart of
At 520, the search application 142 may determine one or more attributes for the user. For example, the user may have one or more of the following attributes: a keyword, a manager, a department, a team, an organization, and the like. Those attributes may be included on a list of attributes.
At 530, the search application 142 searches for other users with similar attributes. For example, search application 142 may search address books, employee directories, organizational information, and the like for other users with similar attributes. Referring to the above example, if the user is Nancy (
At 550, the search application 142 creates a query including the keyword and additional query terms including attributes. For example, search application 142 combines the search term with the list of users that may have “edited”, “authored” or “stored” the document. At 560, the search application 142 performs a keyword search, which is limited to include attributes. For example, if Nancy is performing a search for “Project 1 Schedule,” the search application 142 may create a query that when performed limits the search to results including at least one of the following attributes: “Created by Nancy,” “Created by Paul,” “Created by Karen,” or “Created by Joe.” In the previous example, Nancy's search for “Project 1 Schedule” is limited based on context information, namely documents created by people in her department.
At 570, search application 142 provides the search results to user interface 120, so that a user may view the results.
Although
Moreover, although the above describes the context information as an organizational relationship corresponding to the departments associated with the user performing the search and others associated with the documents being searched, other relationships may be used as well. For example, an organizational relationship may also be implemented that uses a relationship between the user performing the search and a customer, a supplier, or other entities outside the organization.
In some implementations, a user may provide a user identifier and/or a password to login to a computer providing access to one or more documents being searched. Those documents may be copied from their primary storage location to a storage location associated with search application 142 (configured as a search engine). In addition, the context information including organizational information associated with the user and the authors of the documents being searched may be stored in a relational database. The stored context information (i.e., stored in the relational database) may then be copied to the storage location associated with application 142. When a user performs a search using the search application 142, it accesses the associated storage location to search the documents and use the context information.
As used herein, an author of a document includes a creator of the document, a modifier of the document, an owner of a document, a custodian of a document, and any other person identified in metadata associated with the document.
The systems and methods disclosed herein may be embodied in various forms including, for example, a data processor, such as a computer that also includes a database, digital electronic circuitry, firmware, software, or in combinations of them. Moreover, the above-noted features and other aspects and principles of the present disclosed embodiments may be implemented in various environments. Such environments and related applications may be specially constructed for performing the various processes and operations according to the disclosed embodiments or they may include a general-purpose computer or computing platform selectively activated or reconfigured by code to provide the necessary functionality. The processes disclosed herein are not inherently related to any particular computer, network, architecture, environment, or other apparatus, and may be implemented by a suitable combination of hardware, software, and/or firmware. For example, various general-purpose machines may be used with programs written in accordance with teachings of the disclosed embodiments, or it may be more convenient to construct a specialized apparatus or system to perform the required methods and techniques.
The systems and methods disclosed herein may be implemented as a computer program product, i.e., a computer program tangibly embodied in an information carrier, e.g., in a machine readable storage device or in a propagated signal, for execution by, or to control the operation of, data processing apparatus, e.g., a programmable processor, a computer, or multiple computers. A computer program can be written in any form of programming language, including compiled or interpreted languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A computer program can be deployed to be executed on one computer or on multiple computers at one site or distributed across multiple sites and interconnected by a communication network.
The foregoing description is intended to illustrate but not to limit the scope of the invention, which is defined by the scope of the appended claims. Other embodiments are within the scope of the following claims.
