Data processing: database and file management or data structures – Database design – Data structure types
Reexamination Certificate
2000-01-26
2003-11-11
Vu, Kim (Department: 2172)
Data processing: database and file management or data structures
Database design
Data structure types
C707S793000, C707S793000, C707S793000
Reexamination Certificate
active
06647381
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates generally to a method of partitioning and reorganizing physical domains into logical domains, and more particularly to a method of utilizing logical domains for the construction of multi-granular and topic-focused site maps of part of a physical domain or a search space.
2. Description of the Related Art
Use of the internet, and in particular the World Wide Web (WWW or Web), has recently been increasing at a rapid rate. The explosive popularity of the WWW has been accompanied by a tremendous growth in the size of the Web and the scope of its content. Due to the ever-increasing size and complexity of the search space, many queries of the vast WWW, for example, yield such a large number of matched Web pages that the results returned by the search engine are not useful. Even when specific information related to a definite topic is sought, Web users often encounter difficulties in foraging for relevant pages. Many of these difficulties are rooted in the structure of the search space and can be attributed to the deficiencies inherent in the way conventional search techniques and result organization schemes operate.
Generally, the Web includes many Hyper-Text Markup Language (or HTML) documents, or pages, and each page is assigned a unique Universal Resource Locator (URL) for identification and location purposes. The URLs are organized into physical domains; each physical domain is defined as a set of pages associated with a single host, and each page located within a particular physical domain contains the host name in its URL. For example, the URLs www.ccrl.com and www.ccrl.com/dl99ws/ identify individual pages which are both hosted by a Web server (or series of Web servers) having a unique host name (i.e.: www.ccrl.com). The pages identified by these two URLs are, therefore, each in the same physical domain. Since a particular URL represents a unique identifier for every HTML page, the URL is the preferred means utilized by conventional search engines and query processing methods for organizing Web query results in a physical domain.
With many conventional Web search engines, for example, query results or information reported in response to a request are grouped exclusively by physical domain and are presented in the form of a set of clusters of URLs within a particular physical domain. This organizational strategy is advantageous to the extent that the clusters can potentially provide a user with a visualization of the topology of the search space, i.e., how the pages are linked together. A user may thus first locate the most relevant site and browse through matched pages within that site. Organizing query results exclusively by physical domain has two significant limitations, however, especially when a physical domain contains large Web sites.
First, large Web sites tend to contain many matching pages arranged in only a few large, flat-structured, and unorganized clusters. This phenomenon is attributable to the fact that many pages, by virtue of their presence in the same physical domain, have related or similar URLs. For example, many large Web sites, such as Geocities, AOL, and NEC BIGLOBE, are either Internet Service Provider (ISP) sites or Web site hosting providers; consequently, these sites represent enormous physical domains. Typically, if one page contains pertinent information, many pages with similar URLs will also be returned as a match in response to a request for information. Many of the matching pages are only relevant to the extent that their URLs contain a given string of characters, namely, the DNS name; the actual contents of the pages may be totally irrelevant. The inclusion of irrelevant material in the search results shifts the burden of distinguishing between relevant pages and irrelevant pages from the search engine to the user.
Second, even assuming that all of the information returned by the search engine were relevant, grouping results by physical domain does not provide a well organized and convenient way for users to locate the most relevant pages in Web sites. For example, given a query containing the keyword “XML,” many portal sites specializing in XML material, such as www.xml.org and www.w3c.org, tend to offer a large number of matches which, when displayed in the form of a query result, are not categorized or otherwise summarized by a typical search engine. A method of presenting a user with an hierarchical display representing how the hundreds of pages are related in addition to their URL similarities will usually be of greater utility to a user than merely displaying, in list form, hundreds of pages within a given physical domain without any indication as to the way the pages are related.
In addition, while Web site maps can play an essential role in assisting users in navigating a Web site, many site maps can also prove to be inefficient or wholly ineffective with respect to the goal of assisting navigation. Ideally, such maps should provide users with a view of both the contents (i.e.: pages) as well as the link structure (i.e.: topology) of the Web sites they represent. Since, as noted above, the state of the art involves organizing query results and requests for information mainly according to physical domain, the typical site map necessarily reflects only the content and structure of a physical domain. To the extent that the organization of a physical domain is deficient or renders navigation tedious, the site map representing that physical domain can offer little or no assistance to a user interested in finding information regarding a particular topic of interest. A method of organizing results responsive to keyword queries or other requests for information into convenient and usable form should be, therefore, adapted for the construction of site maps having greater utility to users.
Many Web-masters presume that users of their sites have different hardware capabilities, network bandwidths, and preferences for interacting with the site. To support a more user-friendly and pleasant Web surfing experience, many Web sites support several variations with respect to the way information is presented to a user, such as, for example, text modes, graphics modes which may or may not support frames, Java™ scripts, and so forth. Although users with different hardware or bandwidth capabilities are supported, the fact that different users may be browsing for different topics of interest is usually overlooked. For instance, most of the site maps are static, i.e.: predetermined and unalterable. The static nature of the typical site map is most evident at big portal sites that present vast amounts of information covering many diverse subjects. Such a static approach to site mapping is deficient to the extent that it assumes that a single map is suitable for all users who may visit the site.
For example, different users generally have different preferences with respect to visiting a Web site, and different users may visit the same site for different purposes. On any given Web site, for instance, one user may be hunting for particular information, while another user may simply be surfing the Web for enjoyment without any well-defined target in mind. Obviously, the relative expectations of these two users with respect to site map complexity are necessarily different. The former may want to see a detailed map which aids in speedy navigation to a specific directory containing the specific information sought, while the latter may prefer a more abstract map which merely offers a general overview of the contents of the Web site. It is desirable, therefore, to construct site maps which support multiple levels of granularity. A multi-granular site map enables a user selectively to examine different portions of the site in varying degrees of detail, from the very general to the very specific.
As another example, different users generally have different topics of interests in mind when conducting a search. In the case of an on-line “super-store” having many different types of items for sal
Kolak Okan
Li Wen-Syan
Vu Quoc
NEC USA Inc.
Sughrue & Mion, PLLC
To Baoquoc N
Vu Kim
LandOfFree
Method of defining and utilizing logical domains to... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of defining and utilizing logical domains to..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of defining and utilizing logical domains to... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3170685