• 목록
• 답변
• 글쓰기
• 게시판 리스트 옵션
  • 수정
  • 삭제

For example and in one embodiment, if for a question, the deep link search area returns and ranks outcomes for motion pictures increased than different results, the blender/ranker 806 would boost the scores of the outcomes from the media search area 808 B. As another instance and embodiment, if for the query, the deep hyperlink search area returns and ranks outcomes for website indexing online encyclopedia entries larger than different outcomes, the blender/ranker 806 would increase the scores of the results from the wiki search domain 808 C. Using one search area to rank outcomes from different search domains is further described in the FIGS. 500 begins by sending a question to a deep link indexing search service at block 502 . 1000 decreases the ranking rating for a outcome that corresponds to a search domain that has a lower ranked search domain type. Process seven-hundred provides the span rating to the object score at block 716 . 500 in addition to launching the application, course of 500 presents a breadcrumb consumer interface object that signifies to the person methods to return to the set of results displayed at block 506 .

Google Cache and Google Pagespeed (Google Cache shall be displayed in Google Chrome browser). These adverts appear like mini browser home windows and may appear with out the handle bar at the highest. At the top of the page, you’ll see a dialog box with information about when the URL was final cached in the index. 104 features a obtain seed object module 1102 , retrieve object module 1104 , analyze object module 1106 , index deep hyperlink module 1108 , and add additional objects module 1110 . On this embodiment, course of 500 would install the appliance and launch the appliance utilizing the chosen deep hyperlink. With the ranked search domain varieties, process 900 ranks the first set of results using the ranked search domain varieties. 600 ranks the set of results using the consequence scores decided above. 900 returns the ranked first set of results. 900 sends the query to a broad-base search domain, where the broad-base search area covers many different types of doable outcomes. 900 sends the query to the maps, website indexing media, wiki, sites, and/or different search domains, where each of the search domains determines a set of results for the query, corresponding to search domain 808 A-F as described in FIG. Eight above.

806 makes use of the kinds of results from the deep link search area 808 G to rank the outcomes from the other search domains 808 A-F. FIG. Four is a circulation diagram of 1 embodiment of a course of 400 to perform a question search using a deep hyperlink index. If the object does embrace references to other objects that may be crawled, process 300 adds the references to these objects in the group of objects. 1102 obtain the references for the seed objects as described in FIG. 3 , block 302 above. One thousand begins by receiving the first and second set of results at block 1002 . 900 receives a primary set of results from the multiple search domains and a second set of outcomes from the broad-base search area. 806 receives the outcomes from the a number of search domains 808 A-G and ranks these outcomes. 802 features a blender/ranker 806 , and a number of search domains 808 A-G. FIG. 8 is a block diagram of one embodiment of a federator 802 that performs a multi-domain search using results from one search area to rank results from different search domains. 804 E is a set of other search domains that may be accessed by the federator 802 (e.g., a news search domain).

804 B is a search domain associated to media. Process one thousand ranks the search domain sorts utilizing rating scores at block 1008 . 1000 ranks the first set of outcomes utilizing the ranked search area sorts. 900 performs course of a thousand to rank types of broad-base search domain results and use the ranked broad-base varieties to rank results from other search domains. FIG. 7 is a circulation diagram of 1 embodiment of a process seven-hundred to score an object using an ordered quadratic proximity perform. "white truck" would rating the result increased for strings "white Ford truck," "white slightly used truck," however rating decrease "truck with white paint" (out of order) and "white house that features a lawn with a truck" (longer distance between phrases). Current retrieval match phrases contained within textual content or documents in a database. For example and in one embodiment, for the query "TripAdvisor Bardessono," the span with the terms "TripAdvisor-Bardessono" will rating higher because the space between the query phrases is small (one time period) and the match is an so as match. In a single embodiment, process 700 identifies a match by identifying the terms of the query in the article.