Patent US6795820B2 successfully refreshed from Google Patents.
US6795820B2
Expired / Lapsed
US
⏱️ 0.0 years remaining
Metasearch technique that ranks documents obtained from multiple collections
Assignee: Microsoft Corp
•
Filing: 2001-06-20
•
Granted: 2004-09-21
•
Expires: 2021-06-20
Abstract
Abstract In a metasearch conducted across multiple document collections, a multi-phase approach is employed in which local and global statistics are dynamically exchanged between the search engines and the metasearch engine in response to a user's query. In the first phase, the query is transmitted to the search engines from the metasearch engine, and each search engine computes or retrieves previously-computed local statistics for those terms in its associated document collection. In the second phase, each search engine returns its local statistics. A third phase consists of computing metacollection level statistics at the metasearch engine, based upon the information contained in the local statistics. The metacollection level statistics are disseminated to the search engines. In the final phase, the search engines rank the documents in their respective collections pursuant to the metacollection level statistics, and transmit the rankings to the metasearch engine. The metasearch engine merges the results from the individual search engines, to produce a single ranked results list for the user.
Inventors
- 👤 Russell Clark Barnett
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-09 10:53
CPC Classifications
G; G06; G06F; G06F16/00; G06F16/90; G06F16/95; G06F16/951; G06F16/953; G06F16/9532; G06F16/9538; Y; Y10; Y10S; Y10S707/00; Y10S707/99931; Y10S707/99933
Claims Analysis
Total Claims: 136
Claim 1
Independent Claim
194 words
A method for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising the steps of: receiving a query string at a metasearch engine, and transmitting terms in said query to search engines associated with said document collections; at each search engine, dynamically computing local statistics related to said terms for the documents in a collection with which said search engine is associated, including a score normalization factor that comprises a mean document length for the documents in the collection, in response to receipt of said query, and providing said local statistics to the metasearch engine; computing at least one global statistic related to the documents in the metacollection, including a score normalization factor that comprises a mean document length for the documents in the metacollection, in response to receipt of said local statistics at the metasearch engine, and transmitting said global statistic to said search engines; determining relevancy scores for said documents at said search engines in accordance with said global statistic; normalizing said scores in accordance with said normalization factor for the metacollection; and providing references to documents in said metacollection in accordance said relevancy scores.
Claim 2
Independent Claim
194 words
A method for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising the steps of: receiving a query string at a metasearch engine, and transmitting terms in said query to search engines associated with said document collections; at each search engine, dynamically computing local statistics related to said terms for the documents in a collection with which said search engine is associated, including a score normalization factor that comprises a mean document length for the documents in the collection, in response to receipt of said query, and providing said local statistics to the metasearch engine; computing at least one global statistic related to the documents in the metacollection, including a score normalization factor that comprises a mean document length for the documents in the metacollection, in response to receipt of said local statistics at the metasearch engine, and transmitting said global statistic to said search engines; determining relevancy scores for said documents at said search engines in accordance with said global statistic; normalizing said scores in accordance with said normalization factor for the metacollection; and providing references to documents in said metacollection in accordance said relevancy scores.
Claim 3
Dependent Claim
20 words
The method of claim 1 wherein said global statistic is also computed dynamically in response to receipt of said query.
Claim 4
Dependent Claim
26 words
The method of claim 1 wherein the relevancy score for a document is based solely upon the terms in said query which appear in the document.
Claim 5
Dependent Claim
23 words
The method of claim 1 wherein multiple global statistics are computed at the metasearch engine in response to receipt of said local statistics.
Claim 6
Dependent Claim
21 words
The method of claim 4 wherein all global statistics required to determine the relevancy scores are computed at the metasearch engine.
Claim 7
Dependent Claim
28 words
The method of claim 4 wherein a further global statistic is computed at the search engines in response to receipt of multiple global statistics from said metasearch engine.
Claim 8
Dependent Claim
55 words
The method of claim 4 wherein said local statistics include the size of a collection and a measure of the frequency of a given term within the collection, and wherein said global statistics include a measure of the size of the metacollection and a measure of the frequency of a given term within the metacollection.
Claim 9
Dependent Claim
89 words
The method of claim 7 wherein said measure of the size of the collection comprises the number of documents in the collection, said measure of the frequency of a given term within the collection comprises the number of documents in the collection that contain said given term, said measure of the size of the metacollection comprises the number of documents in the metacollection, and said measure of the frequency of a given term within the metacollection comprises the number of documents in the metacollection that contain said given term.
Claim 10
Dependent Claim
87 words
The method of claim 7 wherein said measure of the size of the collection comprises the number of terms in the collection, said measure of the frequency of a given term within the collection comprises the number of occurrences in the collection of said given term, said measure of the size of the metacollection comprises the number of terms in the metacollection, and said measure of the frequency of a given term within the metacollection comprises the number of occurrences in the metacollection of said given term.
Claim 11
Dependent Claim
33 words
The method of claim 7 wherein another global statistic comprises an inverse frequency factor that is computed from the size of the metacollection and the frequency of a given term within the metacollection.
Claim 12
Dependent Claim
16 words
The method of claim 10 wherein said inverse frequency factor is computed at the metasearch engine.
Claim 13
Dependent Claim
18 words
The method of claim 10 wherein said inverse frequency factor is computed at each of the search engines.
Claim 14
Dependent Claim
41 words
The method of claim 1 including the further step of filtering the documents in a collection in accordance with a Boolean condition, and wherein the step of computing local statistics is conducted on the documents which result from said filtering step.
Claim 15
Dependent Claim
26 words
The method of claim 13 wherein the documents in a collection are evaluated in accordance with said Boolean condition in response to receipt of a query.
Claim 16
Dependent Claim
19 words
The method of claim 14 wherein the Boolean condition is specified by a user in conjunction with a query.
Claim 17
Dependent Claim
51 words
The method of claim 14 wherein the query is generated by a user who has access to only a portion of the documents in a collection, and wherein each document in the collection is evaluated in accordance with said Boolean condition regardless of whether the user has access to the document.
Claim 18
Dependent Claim
137 words
The method of claim 1 including the further step at each search engine of sorting references to the documents in a collection into an order determined by one or more criteria, and providing said scores to said metasearch engine in the following steps: a) transmitting a request from said metasearch engine to said search engines for at least one document reference; and b) returning from each search engine a reference to at least the highest-ranking document reference in the sorted order within said search engine; and wherein the step of providing references to documents across said metacollection comprises the following steps: c) sorting the document references returned from said search engines into a metacollection-level order determined by said criteria; and d) providing to a requestor a reference to at least the highest-ranking document in the metacollection-level order.
Claim 19
Dependent Claim
86 words
The method of claim 17 , further including the steps of: e) identifying the search engine that provided said highest-ranking document reference in step (d); f) sending a request from said metasearch engine to the search engine identified in step (e) for at least one additional document reference; g) merging the document returned in response to step (f) into said metacollection-level sorted order; and h) providing to a requestor at least the next-highest-ranking document in the metacollection-level order after the document reference provided in step (d).
Claim 20
Dependent Claim
79 words
The method of claim 17 wherein step (b) comprises returning from said search engine to said metasearch engine a reference to each of the at least two highest-ranking documents in the sorted order within said search engine and storing said references in memory at the metasearch engine; and further including the step of providing to a requestor the next-highest-ranking document in the metacollection-level order after the document reference provided in step (d) from said memory at the metasearch engine.
Claim 21
Dependent Claim
102 words
The method of claim 17 , wherein the document references returned at step (b) are stored in memory at the metasearch engine; and further including the following steps, for each said search engine: e) determining whether said memory contains at least the next-highest-ranking document in said search engine's sorted order; f) if said determination is negative, sending a request from said metasearch engine to said search engine for at least the next-highest-ranking document in the search engine's sorted order; and g) providing to a requestor at least the next-highest-ranking document in the metacollection-level order after the document reference provided in step (d).
Claim 22
Dependent Claim
84 words
The method of claim 17 , further including the steps of: e) sending a request from said metasearch engine to at least one of said search engines for at least one additional document reference; f) repeating at said search engine said step of determining relevancy scores for at least those documents not yet returned by the search engine to the metasearch engine; and g) sending at least one additional document reference from said search engine to said metasearch engine in response to said request.
Claim 23
Dependent Claim
112 words
The method of claim 17 , wherein said step of determining relevancy scores includes the step of storing in memory at each said search engine at least one of said relevancy scores, and further including the steps of: e) sending a request from said metasearch engine to at least one of said search engines for at least one additional document reference; f) at said search engine, determining whether said memory contains sufficient information to respond to said request; g) repeating said step of determining relevancy scores at said search engine; and h) sending at least one additional document reference from said search engine to said metasearch engine in response to said request.
Claim 24
Dependent Claim
16 words
The method of claim 17 wherein each document reference comprises a copy of the associated document.
Claim 25
Dependent Claim
19 words
The method of claim 17 wherein each document reference comprises a link to the location of the associated document.
Claim 26
Dependent Claim
16 words
The method of claim 24 wherein said document reference further includes a description of the document.
Claim 27
Dependent Claim
16 words
The method of claim 17 wherein each document reference comprises an identification of the associated document.
Claim 28
Dependent Claim
12 words
The method of claim 17 wherein said criteria include said relevancy scores.
Claim 29
Dependent Claim
12 words
The method of claim 17 wherein said criteria include an alphabetical ordering.
Claim 30
Dependent Claim
15 words
The method of claim 17 wherein said criteria include a date associated with each document.
Claim 31
Dependent Claim
269 words
The method of claim 1 including the further step of sorting references to the documents in a collection into an order determined by at least one sort criterion, and providing said scores to said metasearch engine according to the following steps: receiving a request at said metasearch engine for M document references; sending a request from said metasearch engine to each of said search engines for K document references; returning from each said search engine to said metasearch engine a reference to each of the K highest ranking document references in the sorted order within said search engine; sorting the document references returned from said search engines into a metacollection-level order determined by said criteria and storing said references in memory at the metasearch engine; providing to a requestor a reference to at least the highest-ranking document in the sorted order within said metasearch engine; receiving at the metasearch engine a subsequent request for N document references; determining, for each search engine, whether references to the documents at said search engine with ranks M+1 through M+N within the metacollection-level order are present in said memory; sending a request for K additional document references from said metasearch engine to each of said search engines for which said determining step is negative; returning from each said search engine a reference to each of the K next-highest-ranking document references in the sorted order within said search engine; storing said references in memory at the metasearch engine; and providing to a requestor a reference to at least the next-highest-ranking document in the metacollection-level order within said metasearch engine that was not provided previously.
Claim 32
Dependent Claim
29 words
The method of claim 1 wherein said step of transmitting terms comprises parsing the query string at the metasearch engine and transmitting each resulting term to said search engines.
Claim 33
Dependent Claim
37 words
The method of claim 1 wherein said step of transmitting terms comprises transmitting the entire query string to said search engines, and said step of computing local statistics includes parsing the query string at each search engine.
Claim 34
Dependent Claim
63 words
The method of claim 32 wherein said step of providing said local statistics to the metasearch engine includes providing a character offset and length for each term within the query string, and said step of computing global statistics includes computing a match between differently parsed terms that are co-located within the query string and subsequently combining global statistics in accordance with said match.
Claim 35
Dependent Claim
24 words
The method of claim 33 wherein said step of computing a match comprises identifying terms that completely enclose one another within said query string.
Claim 36
Independent Claim
236 words
A method for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising the steps of: receiving a query string at a metasearch engine, and transmitting terms in said query to search engines associated with said document collections; at each search engine, dynamically computing local statistics related to said terms for the documents in a collection with which said search engine is associated, including a score normalization factor for the collection, in response to receipt of said query, and providing said local statistics to the metasearch engine, wherein said normalization factor for a collection comprises a local mean value for the number of times the most frequent term in a document appears in that document for each of the documents in the collection; computing at least one global statistic related to the documents in the metacollection, including a score normalization factor for the metacollection, in response to receipt of said local statistics at the metasearch engine, and transmitting said global statistic to said search engines, wherein the normalization factor for the metacollection comprises a mean value for said local mean values across all of the collections in the metacollection; determining relevancy scores for said documents at said search engines in accordance with said global statistic; normalizing said scores in accordance with said normalization factor for the metacollection; and providing references to documents in said metacollection in accordance said relevancy scores.
Claim 37
Independent Claim
236 words
A method for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising the steps of: receiving a query string at a metasearch engine, and transmitting terms in said query to search engines associated with said document collections; at each search engine, dynamically computing local statistics related to said terms for the documents in a collection with which said search engine is associated, including a score normalization factor for the collection, in response to receipt of said query, and providing said local statistics to the metasearch engine, wherein said normalization factor for a collection comprises a local mean value for the number of times the most frequent term in a document appears in that document for each of the documents in the collection; computing at least one global statistic related to the documents in the metacollection, including a score normalization factor for the metacollection, in response to receipt of said local statistics at the metasearch engine, and transmitting said global statistic to said search engines, wherein the normalization factor for the metacollection comprises a mean value for said local mean values across all of the collections in the metacollection; determining relevancy scores for said documents at said search engines in accordance with said global statistic; normalizing said scores in accordance with said normalization factor for the metacollection; and providing references to documents in said metacollection in accordance said relevancy scores.
Claim 38
Independent Claim
230 words
A system for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising: a plurality of local search engines, each being associated with at least one of said document collections, for receiving query terms and computing local statistics related to said terms for the documents in a collection with which said search engine is associated; a plurality of intermediate level search engines, each of which receives said local statistics from a subset of said local search engines, and computes intermediate-level statistics that are based upon the collections of documents associated with the corresponding subset of local search engines; and a metasearch engine that receives the intermediate-level statistics from said plurality of intermediate level search engines and computes global statistics related to the documents in the metacollection, in response to receipt of said intermediate-level statistics at the metasearch engine, and transmits said global statistics to said local search engines; wherein said local search engines determine relevancy scores for the documents in an associated collection in accordance with said global statistics, and provide said scores to the corresponding intermediate level engine, the intermediate level engines combine the scores from the corresponding subset of local search engines and provide the combined scores to the metasearch engine, and the metasearch engine ranks the documents across said metacollection in accordance with relevancy scores received from said intermediate level engines.
Claim 39
Independent Claim
230 words
A system for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising: a plurality of local search engines, each being associated with at least one of said document collections, for receiving query terms and computing local statistics related to said terms for the documents in a collection with which said search engine is associated; a plurality of intermediate level search engines, each of which receives said local statistics from a subset of said local search engines, and computes intermediate-level statistics that are based upon the collections of documents associated with the corresponding subset of local search engines; and a metasearch engine that receives the intermediate-level statistics from said plurality of intermediate level search engines and computes global statistics related to the documents in the metacollection, in response to receipt of said intermediate-level statistics at the metasearch engine, and transmits said global statistics to said local search engines; wherein said local search engines determine relevancy scores for the documents in an associated collection in accordance with said global statistics, and provide said scores to the corresponding intermediate level engine, the intermediate level engines combine the scores from the corresponding subset of local search engines and provide the combined scores to the metasearch engine, and the metasearch engine ranks the documents across said metacollection in accordance with relevancy scores received from said intermediate level engines.
Claim 40
Dependent Claim
37 words
The system of claim 36 wherein each local search engine filters the documents in a collection in accordance with a Boolean condition, and the local statistics are computed for the documents which result from said filtering step.
Claim 41
Dependent Claim
55 words
The system of claim 36 wherein said local statistics include the size of a collection and a measure of the frequency of a given term within the collection, and wherein said global statistics include a measure of the size of the metacollection and a measure of the frequency of a given term within the metacollection.
Claim 42
Dependent Claim
33 words
The system of claim 38 wherein another global statistic comprises an inverse frequency factor that is computed from the size of the metacollection and the frequency of a given term within the metacollection.
Claim 43
Dependent Claim
45 words
The system of claim 38 wherein said local statistics include a score normalization factor for a collection, and the global statistic comprises a score normalization factor for the metacollection, and said metasearch engine normalizes said scores in accordance with said normalization factor for the metacollection.
Claim 44
Dependent Claim
47 words
The system of claim 36 wherein each local search engine sorts the documents in a collection into an order determined by one or more criteria, and transmits said scores to said intermediate engines in subsets that are determined in accordance with the sorted order of the documents.
Claim 45
Dependent Claim
12 words
The system of claim 41 wherein said criteria include said relevancy scores.
Claim 46
Dependent Claim
12 words
The system of claim 41 wherein said criteria include an alphabetical ordering.
Claim 47
Dependent Claim
36 words
The system of claim 41 wherein said intermediate engines merge and sort the subsets of scores received from their associated local search engines, and transmit a further subset of the sorted scores to the metasearch engine.
Claim 48
Independent Claim
242 words
A method for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising the steps of: receiving query terms at a plurality of local search engines, each being associated with at least one of said document collections, and computing local statistics related to said terms for the documents in a collection with which said search engine is associated; transmitting local statistics from subsets of said local search engines to respective ones of a plurality of intermediate level search engines, and computing intermediate-level statistics at each of said intermediate level search engines that are based upon the collections of documents associated with a corresponding subset of local search engines; transmitting the intermediate-level statistics from said plurality of intermediate level search engines to a metasearch engine, computing global statistics related to the documents in the metacollection, in response to receipt of said intermediate-level statistics at the metasearch engine, and transmits said global statistics to said local search engines; determining relevancy scores for the documents in an associated collection in accordance with said global statistics at each of said local search engines, and providing said scores to the corresponding intermediate level engine; combining the scores from the corresponding subset of local search engines at the intermediate level engines and providing the combined scores to the metasearch engine; and ranking the documents across said metacollection at the metasearch engine in accordance with relevancy scores received from said intermediate level engines.
Claim 49
Independent Claim
242 words
A method for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising the steps of: receiving query terms at a plurality of local search engines, each being associated with at least one of said document collections, and computing local statistics related to said terms for the documents in a collection with which said search engine is associated; transmitting local statistics from subsets of said local search engines to respective ones of a plurality of intermediate level search engines, and computing intermediate-level statistics at each of said intermediate level search engines that are based upon the collections of documents associated with a corresponding subset of local search engines; transmitting the intermediate-level statistics from said plurality of intermediate level search engines to a metasearch engine, computing global statistics related to the documents in the metacollection, in response to receipt of said intermediate-level statistics at the metasearch engine, and transmits said global statistics to said local search engines; determining relevancy scores for the documents in an associated collection in accordance with said global statistics at each of said local search engines, and providing said scores to the corresponding intermediate level engine; combining the scores from the corresponding subset of local search engines at the intermediate level engines and providing the combined scores to the metasearch engine; and ranking the documents across said metacollection at the metasearch engine in accordance with relevancy scores received from said intermediate level engines.
Claim 50
Independent Claim
220 words
A system for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising: a plurality of local search engines, each being associated with at least one of said document collections, for receiving query terms and dynamically computing local statistics related to said terms for the documents in a collection with which said search engine is associated in response to receipt of a query, said local statistics including a score normalization factor that comprises a mean document length for the documents in the associated collection; and a metasearch engine which receives said local statistics from said local search engines, and computes global statistics related to the documents in the metacollection, in response to receipt of said local statistics at the metasearch engine, and transmits said global statistics to said local search engines, said global statistics including a score normalization factor that comprises a mean document length for the documents in the metacollection; wherein said local search engines determine relevancy scores for the documents in an associated collection in accordance with said global statistics, normalize said scores in accordance with said normalization factor for the metacollection, and provide said normalized scores to the metasearch engine, and the metasearch engine ranks the documents across said metacollection in accordance with relevancy scores received from said local search engines.
Claim 51
Independent Claim
220 words
A system for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising: a plurality of local search engines, each being associated with at least one of said document collections, for receiving query terms and dynamically computing local statistics related to said terms for the documents in a collection with which said search engine is associated in response to receipt of a query, said local statistics including a score normalization factor that comprises a mean document length for the documents in the associated collection; and a metasearch engine which receives said local statistics from said local search engines, and computes global statistics related to the documents in the metacollection, in response to receipt of said local statistics at the metasearch engine, and transmits said global statistics to said local search engines, said global statistics including a score normalization factor that comprises a mean document length for the documents in the metacollection; wherein said local search engines determine relevancy scores for the documents in an associated collection in accordance with said global statistics, normalize said scores in accordance with said normalization factor for the metacollection, and provide said normalized scores to the metasearch engine, and the metasearch engine ranks the documents across said metacollection in accordance with relevancy scores received from said local search engines.
Claim 52
Dependent Claim
262 words
The system of claim 46 wherein the local search engines sort references to the documents in a collection into an order determined by at least one sort criterion, and provide said scores to said metasearch engine according to the following steps: receiving a request from said metasearch engine for K document references; and returning to said metasearch engine a reference to each of the K highest ranking document references in the sorted order within a local search engine; and wherein said metasearch engine responds to the returned references as follows: sorting the document references returned from said local search engines into a metacollection-level order determined by said criteria and storing said references in memory; providing to a requestor a reference to at least the highest-ranking document in the sorted order within said metasearch engine in response to a request for M document references; in response to a subsequent request for N document references, determining, for each local search engine, whether references to the documents at said local search engine with ranks M+1 through M+N within the metacollection-level order are present in said memory; sending a request for K additional document references to each of said local search engines for which said determining step is negative; receiving from each said local search engine a reference to each of the K next-highest-ranking document references in the sorted order within said local search engine; storing said references in memory; and providing to a requestor a reference to at least the next-highest-ranking document in the metacollection-level order within said metasearch engine that was not provided previously.
Claim 53
Dependent Claim
23 words
The system of claim 46 wherein said metasearch engine parses the query string and transmits each resulting term to said local search engines.
Claim 54
Dependent Claim
26 words
The system of claim 46 wherein the entire query string is transmitted to said local search engines, and said local search engines parse the query string.
Claim 55
Dependent Claim
55 words
The system of claim 49 wherein said local search engines provide a character offset and length for each term within the query string to the metasearch engine, and said metasearch engine computes a match between differently parsed terms that are co-located within the query string and subsequently combines global statistics in accordance with said match.
Claim 56
Dependent Claim
20 words
The system of claim 50 wherein said metasearch engine identifies terms that completely enclose one another within said query string.
Claim 57
Dependent Claim
21 words
The system of claim 46 wherein said metasearch engine computes said global statistics dynamically in response to receipt of said query.
Claim 58
Dependent Claim
26 words
The system of claim 46 wherein the relevancy scores for documents are based solely upon the terms in said query which appear in the respective documents.
Claim 59
Dependent Claim
21 words
The system of claim 46 wherein all global statistics required to determine the relevancy scores are computed at the metasearch engine.
Claim 60
Dependent Claim
28 words
The system of claim 46 wherein a further global statistic is computed at the search engines in response to receipt of said global statistics from said metasearch engine.
Claim 61
Dependent Claim
38 words
The system of claim 46 wherein said local search engines filter the documents in an associated collection in accordance with a Boolean condition, and wherein the local statistics are computed on the documents which result from said filtering.
Claim 62
Dependent Claim
26 words
The system of claim 56 wherein the documents in a collection are evaluated in accordance with said Boolean condition in response to receipt of a query.
Claim 63
Dependent Claim
19 words
The system of claim 57 wherein the Boolean condition is specified by a user in conjunction with a query.
Claim 64
Dependent Claim
51 words
The system of claim 57 wherein the query is generated by a user who has access to only a portion of the documents in a collection, and wherein each document in the collection is evaluated in accordance with said Boolean condition regardless of whether the user has access to the document.
Claim 65
Dependent Claim
55 words
The system of claim 46 wherein said local statistics include the size of a collection and a measure of the frequency of a given term within the collection, and wherein said global statistics include a measure of the size of the metacollection and a measure of the frequency of a given term within the metacollection.
Claim 66
Dependent Claim
89 words
The system of claim 60 wherein said measure of the size of the collection comprises the number of documents in the collection, said measure of the frequency of a given term within the collection comprises the number of documents in the collection that contain said given term, said measure of the size of the metacollection comprises the number of documents in the metacollection, and said measure of the frequency of a given term within the metacollection comprises the number of documents in the metacollection that contain said given term.
Claim 67
Dependent Claim
87 words
The system of claim 60 wherein said measure of the size of the collection comprises the number of terms in the collection, said measure of the frequency of a given term within the collection comprises the number of occurrences in the collection of said given term, said measure of the size of the metacollection comprises the number of terms in the metacollection, and said measure of the frequency of a given term within the metacollection comprises the number of occurrences in the metacollection of said given term.
Claim 68
Dependent Claim
33 words
The system of claim 60 wherein another global statistic comprises an inverse frequency factor that is computed from the size of the metacollection and the frequency of a given term within the metacollection.
Claim 69
Dependent Claim
16 words
The system of claim 63 wherein said inverse frequency factor is computed at the metasearch engine.
Claim 70
Dependent Claim
19 words
The system of claim 63 wherein said inverse frequency factor is computed at each of the local search engines.
Claim 71
Dependent Claim
121 words
The system of claim 46 wherein each local search engine sorts references to the documents in a collection into an order determined by one or more criteria, and provides said scores to said metasearch engine by: a) receiving a request from said metasearch engine for at least one document reference; and b) returning a reference to at least the highest-ranking document reference in the sorted order within said local search engine; and wherein said metasearch engine provides references to documents across said metacollection by: c) sorting the document references returned from said local search engines into a metacollection-level order determined by said criteria; and d) providing to a requestor a reference to at least the highest-ranking document in the metacollection-level order.
Claim 72
Dependent Claim
77 words
The system of claim 66 , wherein said metasearch engine further operates to: identify the local search engine that provided said highest-ranking document reference; send a request from said metasearch engine to identified local search engine for at least one additional document reference; merge the document returned in response to said request into said metacollection-level sorted order; and provide to a requestor at least the next-highest-ranking document in the metacollection-level order after the previously provided document reference.
Claim 73
Dependent Claim
59 words
The method of claim 66 wherein each local search engine returns to said metasearch engine a reference to each of the at least two highest-ranking documents in the sorted order within said search engine; and wherein the metasearch engine stores said references in memory and provides to a requestor the next-highest-ranking document in the metacollection-level order from said memory.
Claim 74
Dependent Claim
101 words
The system of claim 66 , wherein the document references returned from the local search engines are stored in memory at the metasearch engine; and wherein said metasearch engine performs the following steps, for each local search engine: determining whether said memory contains at least the next-highest-ranking document in said local search engine's sorted order; if said determination is negative, sending a request to said local search engine for at least the next-highest-ranking document in the local search engine's sorted order; and providing to a requestor at least the next-highest-ranking document in the metacollection-level order after the previously provided document reference.
Claim 75
Dependent Claim
73 words
The system of claim 66 , wherein said metasearch engine sends a request to at least one of said local search engines for at least one additional document reference; and in response thereto a local search engine determines relevancy scores for at least those documents not yet returned by the search engine to the metasearch engine, and sends at least one additional document reference to said metasearch engine in response to said request.
Claim 76
Dependent Claim
84 words
The method of claim 66 , wherein each local search engine stores in memory at least one of said relevancy scores, and wherein said metasearch engine sends a request to at least one of said local search engines for at least one additional document reference; said local search engine determines whether said memory contains sufficient information to respond to said request, and if not, determines relevancy scores and sends at least one additional document reference to said metasearch engine in response to said request.
Claim 77
Dependent Claim
16 words
The system of claim 66 wherein each document reference comprises a copy of the associated document.
Claim 78
Dependent Claim
19 words
The system of claim 66 wherein each document reference comprises a link to the location of the associated document.
Claim 79
Dependent Claim
16 words
The system of claim 73 wherein said document reference further includes a description of the document.
Claim 80
Dependent Claim
16 words
The system of claim 66 wherein each document reference comprises an identification of the associated document.
Claim 81
Dependent Claim
12 words
The system of claim 66 wherein said criteria include said relevancy scores.
Claim 82
Dependent Claim
12 words
The system of claim 66 wherein said criteria include an alphabetical ordering.
Claim 83
Dependent Claim
15 words
The system of claim 66 wherein said criteria include a date associated with each document.
Claim 84
Independent Claim
207 words
A system for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising: associated in response to receipt of a query, said local statistics including a score normalization factor for a collection that comprises a local mean value for the number of times the most frequent term in a document appears in that document for each of the documents in the collection; a metasearch engine which receives said local statistics from said local search engines, and computes global statistics related to the documents in the metacollection, in response to receipt of said local statistics at the metasearch engine, and transmits said global statistics to said local search engines, wherein the global statistics include a score normalization factor for the metacollection that comprises a mean value for said local mean values across all of the collections in the metacollection; wherein said local search engines determine relevancy scores for the documents in an associated collection in accordance with said global statistics, normalize said scores in accordance with said normalization factor for the metacollection and provide said normalized scores to the metasearch engine, and the metasearch engine ranks the documents across said metacollection in accordance with relevancy scores received from said local search engines.
Claim 85
Independent Claim
207 words
A system for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising: associated in response to receipt of a query, said local statistics including a score normalization factor for a collection that comprises a local mean value for the number of times the most frequent term in a document appears in that document for each of the documents in the collection; a metasearch engine which receives said local statistics from said local search engines, and computes global statistics related to the documents in the metacollection, in response to receipt of said local statistics at the metasearch engine, and transmits said global statistics to said local search engines, wherein the global statistics include a score normalization factor for the metacollection that comprises a mean value for said local mean values across all of the collections in the metacollection; wherein said local search engines determine relevancy scores for the documents in an associated collection in accordance with said global statistics, normalize said scores in accordance with said normalization factor for the metacollection and provide said normalized scores to the metasearch engine, and the metasearch engine ranks the documents across said metacollection in accordance with relevancy scores received from said local search engines.
Claim 86
Independent Claim
192 words
A method for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising the steps of: receiving a query string at a metasearch engine, and transmitting terms in said query to search engines associated with said document collections; at each search engine, computing local statistics related to said terms for the documents in a collection with which said search engine is associated, including a score normalization factor for a collection that comprises a mean document length for the documents in the collection, and providing said local statistics to the metasearch engine; computing at least one global statistic related to the documents in the metacollection in response to receipt of said local statistics at the metasearch engine, including a score normalization factor for the metacollection that comprises a mean document length for the documents in the metacollection, and transmitting said global statistic to said search engines; normalizing said scores in accordance with said normalization factor for the metacollection; determining relevancy scores for said documents at said search engines in accordance with said global statistic; and providing references to documents in said metacollection in accordance said relevancy scores.
Claim 87
Independent Claim
192 words
A method for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising the steps of: receiving a query string at a metasearch engine, and transmitting terms in said query to search engines associated with said document collections; at each search engine, computing local statistics related to said terms for the documents in a collection with which said search engine is associated, including a score normalization factor for a collection that comprises a mean document length for the documents in the collection, and providing said local statistics to the metasearch engine; computing at least one global statistic related to the documents in the metacollection in response to receipt of said local statistics at the metasearch engine, including a score normalization factor for the metacollection that comprises a mean document length for the documents in the metacollection, and transmitting said global statistic to said search engines; normalizing said scores in accordance with said normalization factor for the metacollection; determining relevancy scores for said documents at said search engines in accordance with said global statistic; and providing references to documents in said metacollection in accordance said relevancy scores.
Claim 88
Independent Claim
216 words
A method for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising the steps of: receiving a query string at a metasearch engine, and transmitting terms in said query to search engines associated with said document collections; at each search engine, computing local statistics related to said terms for the documents in a collection with which said search engine is associated, including a score normalization factor for a collection that comprises a local mean value for the number of times the most frequent term in a document appears in that document for each of the documents in the collection, and providing said local statistics to the metasearch engine; computing at least one global statistic related to the documents in the metacollection in response to receipt of said local statistics at the metasearch engine, including a score normalization factor for the metacollection that comprises a mean value for said local mean values across all of the collections in the metacollection, and transmitting said global statistic to said search engines; normalizing said scores in accordance with said normalization factor for the metacollection; determining relevancy scores for said documents at said search engines in accordance with said global statistic; and providing references to documents in said metacollection in accordance said relevancy scores.
Claim 89
Independent Claim
216 words
A method for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising the steps of: receiving a query string at a metasearch engine, and transmitting terms in said query to search engines associated with said document collections; at each search engine, computing local statistics related to said terms for the documents in a collection with which said search engine is associated, including a score normalization factor for a collection that comprises a local mean value for the number of times the most frequent term in a document appears in that document for each of the documents in the collection, and providing said local statistics to the metasearch engine; computing at least one global statistic related to the documents in the metacollection in response to receipt of said local statistics at the metasearch engine, including a score normalization factor for the metacollection that comprises a mean value for said local mean values across all of the collections in the metacollection, and transmitting said global statistic to said search engines; normalizing said scores in accordance with said normalization factor for the metacollection; determining relevancy scores for said documents at said search engines in accordance with said global statistic; and providing references to documents in said metacollection in accordance said relevancy scores.
Claim 90
Independent Claim
420 words
A method for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising the steps of: receiving a query string at a metasearch engine, and transmitting terms in said query to search engines associated with said document collections; at each search engine, computing local statistics related to said terms for the documents in a collection with which said search engine is associated, including a score normalization factor that comprises a mean document length for the documents in the collection, and providing said local statistics to the metasearch engine; computing at least one global statistic related to the documents in the metacollection, including a score normalization factor that comprises a mean document length for the documents in the metacollection, in response to receipt of said local statistics at the metasearch engine, and transmitting said global statistic to said search engines; determining relevancy scores for said documents at said search engines in accordance with said global statistic and sorting references to the documents in a collection into an order determined by at least one sort criterion; normalizing said scores in accordance with said normalization factor for the metacollection; receiving a request at said metasearch engine for M document references; sending a request from said metasearch engine to each of said search engines for K document references; returning from each said search engine to said metasearch engine a reference to each of the K highest ranking document references in the sorted order within said search engine; sorting the document references returned from said search engines into a metacollection-level order determined by said criteria and storing said references in memory at the metasearch engine; providing to a requestor a reference to at least the highest-ranking document in the sorted order within said metasearch engine; receiving at the metasearch engine a subsequent request for N document references; determining, for each search engine, whether references to the documents at said search engine with ranks M+1 through M+N within the metacollection-level order are present in said memory; sending a request for K additional document references from said metasearch engine to each of said search engines for which said determining step is negative; returning from each said search engine a reference to each of the K next-highest-ranking document references in the sorted order within said search engine; storing said references in memory at the metasearch engine; and providing to a requestor a reference to at least the next-highest-ranking document in the metacollection-level order within said metasearch engine that was not provided previously.
Claim 91
Independent Claim
420 words
A method for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising the steps of: receiving a query string at a metasearch engine, and transmitting terms in said query to search engines associated with said document collections; at each search engine, computing local statistics related to said terms for the documents in a collection with which said search engine is associated, including a score normalization factor that comprises a mean document length for the documents in the collection, and providing said local statistics to the metasearch engine; computing at least one global statistic related to the documents in the metacollection, including a score normalization factor that comprises a mean document length for the documents in the metacollection, in response to receipt of said local statistics at the metasearch engine, and transmitting said global statistic to said search engines; determining relevancy scores for said documents at said search engines in accordance with said global statistic and sorting references to the documents in a collection into an order determined by at least one sort criterion; normalizing said scores in accordance with said normalization factor for the metacollection; receiving a request at said metasearch engine for M document references; sending a request from said metasearch engine to each of said search engines for K document references; returning from each said search engine to said metasearch engine a reference to each of the K highest ranking document references in the sorted order within said search engine; sorting the document references returned from said search engines into a metacollection-level order determined by said criteria and storing said references in memory at the metasearch engine; providing to a requestor a reference to at least the highest-ranking document in the sorted order within said metasearch engine; receiving at the metasearch engine a subsequent request for N document references; determining, for each search engine, whether references to the documents at said search engine with ranks M+1 through M+N within the metacollection-level order are present in said memory; sending a request for K additional document references from said metasearch engine to each of said search engines for which said determining step is negative; returning from each said search engine a reference to each of the K next-highest-ranking document references in the sorted order within said search engine; storing said references in memory at the metasearch engine; and providing to a requestor a reference to at least the next-highest-ranking document in the metacollection-level order within said metasearch engine that was not provided previously.
Claim 92
Dependent Claim
19 words
The method of claim 82 wherein said global statistic is computed dynamically in response to receipt of said query.
Claim 93
Dependent Claim
20 words
The method of claim 83 wherein said local statistics are also computed dynamically in response to receipt of said query.
Claim 94
Dependent Claim
26 words
The method of claim 82 wherein the relevancy score for a document is based solely upon the terms in said query which appear in the document.
Claim 95
Dependent Claim
23 words
The method of claim 82 wherein multiple global statistics are computed at the metasearch engine in response to receipt of said local statistics.
Claim 96
Dependent Claim
21 words
The method of claim 86 wherein all global statistics required to determine the relevancy scores are computed at the metasearch engine.
Claim 97
Dependent Claim
28 words
The method of claim 86 wherein a further global statistic is computed at the search engines in response to receipt of multiple global statistics from said metasearch engine.
Claim 98
Dependent Claim
55 words
The method of claim 86 wherein said local statistics include the size of a collection and a measure of the frequency of a given term within the collection, and wherein said global statistics include a measure of the size of the metacollection and a measure of the frequency of a given term within the metacollection.
Claim 99
Dependent Claim
89 words
The method of claim 89 wherein said measure of the size of the collection comprises the number of documents in the collection, said measure of the frequency of a given term within the collection comprises the number of documents in the collection that contain said given term, said measure of the size of the metacollection comprises the number of documents in the metacollection, and said measure of the frequency of a given term within the metacollection comprises the number of documents in the metacollection that contain said given term.
Claim 100
Dependent Claim
87 words
The method of claim 89 wherein said measure of the size of the collection comprises the number of terms in the collection, said measure of the frequency of a given term within the collection comprises the number of occurrences in the collection of said given term, said measure of the size of the metacollection comprises the number of terms in the metacollection, and said measure of the frequency of a given term within the metacollection comprises the number of occurrences in the metacollection of said given term.
Claim 101
Dependent Claim
33 words
The method of claim 89 wherein another global statistic comprises an inverse frequency factor that is computed from the size of the metacollection and the frequency of a given term within the metacollection.
Claim 102
Dependent Claim
16 words
The method of claim 92 wherein said inverse frequency factor is computed at the metasearch engine.
Claim 103
Dependent Claim
18 words
The method of claim 92 wherein said inverse frequency factor is computed at each of the search engines.
Claim 104
Dependent Claim
41 words
The method of claim 82 including the further step of filtering the documents in a collection in accordance with a Boolean condition, and wherein the step of computing local statistics is conducted on the documents which result from said filtering step.
Claim 105
Dependent Claim
26 words
The method of claim 95 wherein the documents in a collection are evaluated in accordance with said Boolean condition in response to receipt of a query.
Claim 106
Dependent Claim
19 words
The method of claim 96 wherein the Boolean condition is specified by a user in conjunction with a query.
Claim 107
Dependent Claim
51 words
The method of claim 96 wherein the query is generated by a user who has access to only a portion of the documents in a collection, and wherein each document in the collection is evaluated in accordance with said Boolean condition regardless of whether the user has access to the document.
Claim 108
Dependent Claim
137 words
The method of claim 82 including the further step at each search engine of sorting references to the documents in a collection into an order determined by one or more criteria, and providing said scores to said metasearch engine in the following steps: a) transmitting a request from said metasearch engine to said search engines for at least one document reference; and b) returning from each search engine a reference to at least the highest-ranking document reference in the sorted order within said search engine; and wherein the step of providing references to documents across said metacollection comprises the following steps: c) sorting the document references returned from said search engines into a metacollection-level order determined by said criteria; and d) providing to a requestor a reference to at least the highest-ranking document in the metacollection-level order.
Claim 109
Dependent Claim
86 words
The method of claim 99 , further including the steps of: e) identifying the search engine that provided said highest-ranking document reference in step (d); f) sending a request from said metasearch engine to the search engine identified in step (e) for at least one additional document reference; g) merging the document returned in response to step (f) into said metacollection-level sorted order; and h) providing to a requestor at least the next-highest-ranking document in the metacollection-level order after the document reference provided in step (d).
Claim 110
Dependent Claim
79 words
The method of claim 99 wherein step (b) comprises returning from said search engine to said metasearch engine a reference to each of the at least two highest-ranking documents in the sorted order within said search engine and storing said references in memory at the metasearch engine; and further including the step of providing to a requestor the next-highest-ranking document in the metacollection-level order after the document reference provided in step (d) from said memory at the metasearch engine.
Claim 111
Dependent Claim
102 words
The method of claim 99 , wherein the document references returned at step (b) are stored in memory at the metasearch engine; and further including the following steps, for each said search engine: e) determining whether said memory contains at least the next-highest-ranking document in said search engine's sorted order; f) if said determination is negative, sending a request from said metasearch engine to said search engine for at least the next-highest-ranking document in the search engine's sorted order; and g) providing to a requestor at least the next-highest-ranking document in the metacollection-level order after the document reference provided in step (d).
Claim 112
Dependent Claim
84 words
The method of claim 99 , further including the steps of: e) sending a request from said metasearch engine to at least one of said search engines for at least one additional document reference; f) repeating at said search engine said step of determining relevancy scores for at least those documents not yet returned by the search engine to the metasearch engine; and g) sending at least one additional document reference from said search engine to said metasearch engine in response to said request.
Claim 113
Dependent Claim
112 words
The method of claim 99 , wherein said step of determining relevancy scores includes the step of storing in memory at each said search engine at least one of said relevancy scores, and further including the steps of: e) sending a request from said metasearch engine to at least one of said search engines for at least one additional document reference; f) at said search engine, determining whether said memory contains sufficient information to respond to said request; g) repeating said step of determining relevancy scores at said search engine; and h) sending at least one additional document reference from said search engine to said metasearch engine in response to said request.
Claim 114
Dependent Claim
16 words
The method of claim 99 wherein each document reference comprises a copy of the associated document.
Claim 115
Dependent Claim
19 words
The method of claim 99 wherein each document reference comprises a link to the location of the associated document.
Claim 116
Dependent Claim
16 words
The method of claim 106 wherein said document reference further includes a description of the document.
Claim 117
Dependent Claim
16 words
The method of claim 99 wherein each document reference comprises an identification of the associated document.
Claim 118
Dependent Claim
12 words
The method of claim 99 wherein said criteria include said relevancy scores.
Claim 119
Dependent Claim
12 words
The method of claim 99 wherein said criteria include an alphabetical ordering.
Claim 120
Dependent Claim
15 words
The method of claim 99 wherein said criteria include a date associated with each document.
Claim 121
Dependent Claim
29 words
The method of claim 82 wherein said step of transmitting terms comprises parsing the query string at the metasearch engine and transmitting each resulting term to said search engines.
Claim 122
Dependent Claim
37 words
The method of claim 82 wherein said step of transmitting terms comprises transmitting the entire query string to said search engines, and said step of computing local statistics includes parsing the query string at each search engine.
Claim 123
Dependent Claim
63 words
The method of claim 113 wherein said step of providing said local statistics to the metasearch engine includes providing a character offset and length for each term within the query string, and said step of computing global statistics includes computing a match between differently parsed terms that are co-located within the query string and subsequently combining global statistics in accordance with said match.
Claim 124
Dependent Claim
24 words
The method of claim 114 wherein said step of computing a match comprises identifying terms that completely enclose one another within said query string.
Claim 125
Independent Claim
458 words
A method for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising the steps of: receiving a query string at a metasearch engine, and transmitting terms in said query to search engines associated with said document collections; at each search engine, computing local statistics related to said terms for the documents in a collection with which said search engine is associated, and providing said local statistics to the metasearch engine, wherein said local statistics include a score normalization factor for a collection that comprises a local mean value for the number of times the most frequent term in a document appears in that document for each of the documents in the collection; computing at least one global statistic related to the documents in the metacollection, in response to receipt of said local statistics at the metasearch engine, and transmitting said global statistic to said search engines, wherein the global statistic includes a score normalization factor for the metacollection that comprises a mean value for said local mean values across all of the collections in the metacollection; determining relevancy scores for said documents at said search engines in accordance with said global statistic and sorting references to the documents in a collection into an order determined by at least one sort criterion; normalizing said scores in accordance with said normalization factor for the metacollection; receiving a request at said metasearch engine for M document references; sending a request from said metasearch engine to each of said search engines for K document references; returning from each said search engine to said metasearch engine a reference to each of the K highest ranking document references in the sorted order within said search engine; sorting the document references returned from said search engines into a metacollection-level order determined by said criteria and storing said references in memory at the metasearch engine; providing to a requestor a reference to at least the highest-ranking document in the sorted order within said metasearch engine; receiving at the metasearch engine a subsequent request for N document references; determining, for each search engine, whether references to the documents at said search engine with ranks M+1 through M+N within the metacollection-level order are present in said memory; sending a request for K additional document references from said metasearch engine to each of said search engines for which said determining step is negative; returning from each said search engine a reference to each of the K next-highest-ranking document references in the sorted order within said search engine; storing said references in memory at the metasearch engine; and providing to a requestor a reference to at least the next-highest-ranking document in the metacollection-level order within said metasearch engine that was not provided previously.
Claim 126
Independent Claim
458 words
A method for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising the steps of: receiving a query string at a metasearch engine, and transmitting terms in said query to search engines associated with said document collections; at each search engine, computing local statistics related to said terms for the documents in a collection with which said search engine is associated, and providing said local statistics to the metasearch engine, wherein said local statistics include a score normalization factor for a collection that comprises a local mean value for the number of times the most frequent term in a document appears in that document for each of the documents in the collection; computing at least one global statistic related to the documents in the metacollection, in response to receipt of said local statistics at the metasearch engine, and transmitting said global statistic to said search engines, wherein the global statistic includes a score normalization factor for the metacollection that comprises a mean value for said local mean values across all of the collections in the metacollection; determining relevancy scores for said documents at said search engines in accordance with said global statistic and sorting references to the documents in a collection into an order determined by at least one sort criterion; normalizing said scores in accordance with said normalization factor for the metacollection; receiving a request at said metasearch engine for M document references; sending a request from said metasearch engine to each of said search engines for K document references; returning from each said search engine to said metasearch engine a reference to each of the K highest ranking document references in the sorted order within said search engine; sorting the document references returned from said search engines into a metacollection-level order determined by said criteria and storing said references in memory at the metasearch engine; providing to a requestor a reference to at least the highest-ranking document in the sorted order within said metasearch engine; receiving at the metasearch engine a subsequent request for N document references; determining, for each search engine, whether references to the documents at said search engine with ranks M+1 through M+N within the metacollection-level order are present in said memory; sending a request for K additional document references from said metasearch engine to each of said search engines for which said determining step is negative; returning from each said search engine a reference to each of the K next-highest-ranking document references in the sorted order within said search engine; storing said references in memory at the metasearch engine; and providing to a requestor a reference to at least the next-highest-ranking document in the metacollection-level order within said metasearch engine that was not provided previously.
Claim 127
Independent Claim
220 words
A system for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising: a plurality of local search engines, each being associated with at least one of said document collections, for receiving query terms and computing local statistics related to said terms for the documents in a collection with which said search engine is associated; at least one additional search engine which receives said local statistics from a subset of said local search engines, and computes comprehensive statistics that are based upon the collections of documents associated with the corresponding subset of local search engines; and a further search engine that receives the comprehensive statistics from said additional search engine and statistics from at least one other search engine, and computes global statistics for the documents to which the received statistics pertain, and transmits said global statistics to said local search engines; wherein said local search engines determine relevancy scores for the documents in an associated collection in accordance with said global statistics, and provide said scores to the additional search engine, the additional search engine combines the scores from the corresponding subset of local search engines and provides the combined scores to the further search engine, and the further search engine ranks the documents across said metacollection in accordance with received relevancy scores.
Claim 128
Independent Claim
220 words
A system for identifying and ranking documents contained in a plurality of document collections that form a metacollection, comprising: a plurality of local search engines, each being associated with at least one of said document collections, for receiving query terms and computing local statistics related to said terms for the documents in a collection with which said search engine is associated; at least one additional search engine which receives said local statistics from a subset of said local search engines, and computes comprehensive statistics that are based upon the collections of documents associated with the corresponding subset of local search engines; and a further search engine that receives the comprehensive statistics from said additional search engine and statistics from at least one other search engine, and computes global statistics for the documents to which the received statistics pertain, and transmits said global statistics to said local search engines; wherein said local search engines determine relevancy scores for the documents in an associated collection in accordance with said global statistics, and provide said scores to the additional search engine, the additional search engine combines the scores from the corresponding subset of local search engines and provides the combined scores to the further search engine, and the further search engine ranks the documents across said metacollection in accordance with received relevancy scores.
Claim 129
Dependent Claim
37 words
The system of claim 117 wherein each local search engine filters the documents in a collection in accordance with a Boolean condition, and the local statistics are computed for the documents which result from said filtering step.
Claim 130
Dependent Claim
55 words
The system of claim 117 wherein said local statistics include the size of a collection and a measure of the frequency of a given term within the collection, and wherein said global statistics include a measure of the size of the metacollection and a measure of the frequency of a given term within the metacollection.
Claim 131
Dependent Claim
33 words
The system of claim 119 wherein another global statistic comprises an inverse frequency factor that is computed from the size of the metacollection and the frequency of a given term within the metacollection.
Claim 132
Dependent Claim
46 words
The system of claim 119 wherein said local statistics include a score normalization factor for a collection, and the global statistic comprises a score normalization factor for the metacollection, and said further search engine normalizes said scores in accordance with said normalization factor for the metacollection.
Claim 133
Dependent Claim
48 words
The system of claim 117 wherein each local search engine sorts the documents in a collection into an order determined by one or more criteria, and transmits said scores to said additional search engine in subsets that are determined in accordance with the sorted order of the documents.
Claim 134
Dependent Claim
12 words
The system of claim 122 wherein said criteria include said relevancy scores.
Claim 135
Dependent Claim
12 words
The system of claim 122 wherein said criteria include an alphabetical ordering.
Claim 136
Dependent Claim
37 words
The system of claim 122 wherein said additional search engine merges and sorts the subsets of scores received from said local search engines, and transmits a further subset of the sorted scores to the further search engine.
References Cited (Backward Citations)
US5265244A
Cited
US6029195A
Cited
US5642502A
Cited
US5826261A
Cited
US6070158A
Cited
US5806061A
Cited
US6212517B1
Cited
US6163782A
Cited
US6078924A
Cited
US6553310B1
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2007-09-21 | 2008-03-22 | 2008-09-21 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2011-09-21 | 2012-03-22 | 2012-09-21 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2015-09-21 | 2016-03-22 | 2016-09-21 | $7700.00 | Window Closed / Lapsed |