Method for performing a search of a plurality of documents for similarity to a plurality of query words

Claim 2

A computer implemented method for performing a search of a plurality of documents for similarity to a query having multiple terms using an inverted index of the plurality of documents, the method comprising the sequential steps of: (a) searching the inverted index for each of the query terms and determining a number of occurrences of the query terms in a first document; (b) calculating a similarity score for the first document based on the number of occurrences of the query terms in the first document; (c) searching the inverted index for each of the query terms and determining a number of occurrences of the query terms in a subsequent document; (d) calculating a similarity score for the subsequent document based on the number of occurrences of the query terms in the subsequent document; and (e) repeating steps (c) and (d) until a similarity score has been calculated for each of the plurality of documents.

Claim 3

The method of claim 1, further comprising the step of ranking the plurality of documents based on the calculated similarity scores.

Claim 4

The method of claim 1, further comprising the step of producing a list of the plurality documents that is ordered based on the calculated similarity scores.

Claim 5

The method of claim 1, wherein the query terms comprise images, and wherein the inverted index is an inverted index of images that occur in the plurality of documents.

Claim 6

The method of claim 1, further comprising the step of storing a predetermined number of the highest similarity scores in a memory as the searching and calculating steps are performed.

Claim 7

The method of claim 1, wherein a similarity score for each document is calculated before a search is performed for occurrences of the query terms in a subsequent document.

Claim 8

The method of claim 1, wherein, the steps of calculating a similarity score comprise calculating a similarity score based on a user defined formula and a number of occurrences of the query terms in a document.

Claim 9

A computer implemented method for performing a search of a plurality of documents for similarity to a query having multiple terms using an inverted index of the plurality of documents, comprising the sequential steps of: (a) searching the inverted index for each of the query terms and determining a number of occurrences of the query terms in a first document; (b) calculating a similarity score for the first document based on the number of occurrences of the query terms in the first document; (c) storing the first document's similarity score in a memory; (d) searching the inverted index for each of the query terms and determining a number of occurrences of the query terms in a subsequent document; (e) calculating a similarity score for the subsequent document based on the number of occurrences of the query terms in the subsequent document; (f) storing the subsequent document's similarity score in the memory when fewer than a predetermined number of similarity scores are stored in the memory; (g) comparing the subsequent document's similarity score to the similarity scores stored in the memory, deleting the lowest similarity score stored in the memory, and storing the subsequent document's similarity score in the memory when the subsequent document's similarity score is higher than the lowest similarity score stored in the memory and more than a predetermined number of similarity scores are stored in the memory; and (h) repeating steps (d) through (g) until a similarity score has been calculated for each of the plurality of documents.

Claim 10

A computer implemented method for performing a search of a plurality of documents for similarity to a query having multiple terms using an inverted index of the plurality of documents, comprising the sequential steps of: (a) searching the inverted index for each of the query terms and determining a number of occurrences of the query terms in a first document; (b) calculating a similarity score for the first document based on the number of occurrences of the query terms in the first document; (c) storing the first document's similarity score in a memory; (d) searching the inverted index for each of the query terms and determining a number of occurrences of the query terms in a subsequent document; (e) calculating a similarity score for the subsequent document based on the number of occurrences of the query terms in the subsequent document; (f) storing the subsequent document's similarity score in the memory when fewer than a predetermined number of similarity scores are stored in the memory; (g) comparing the subsequent document's similarity score to the similarity scores stored in the memory, deleting the lowest similarity score stored in the memory, and storing the subsequent document's similarity score in the memory when the subsequent document's similarity score is higher than the lowest similarity score stored in the memory and more than a predetermined number of similarity scores are stored in the memory; and (h) repeating steps (d) through (g) until a similarity score has been calculated for each of the plurality of documents.

Claim 11

The method of claim 8, further comprising the step of ranking the documents whose similarity scores are stored in the memory based on the calculated similarity scores.

Claim 12

The method of claim 8, further comprising the step of producing a list of the documents whose similarity scores are stored in the memory, the list being ordered based on the calculated similarity scores.

Claim 13

The method of claim 8, wherein the query terms comprise images, and wherein the inverted index is an inverted index of images that occur in the plurality of documents.

Claim 14

The method of claim 8, wherein a similarity score for each document is calculated before a search is performed for occurrences of the query terms in a subsequent document.

Claim 15

The method of claim 8, wherein the steps of calculating a similarity score comprise calculating a similarity score based on a user defined formula and a number of occurrences of the query terms in a document.

Claim 16

A computer implemented method for performing a search of a corpus of documents for similarity to a query having multiple terms using an inverted index of the corpus of documents, the method comprising the sequential steps of: (a) retrieving a first portion of the inverted index corresponding to a first plurality of documents from a first memory into a second memory; (b) searching the first portion of the inverted index for each of the query terms and determining a number of occurrences of the query terms in each of the documents in the first portion of the inverted index; (c) calculating similarity scores for each of the documents in the first portion of the inverted index based on the number of occurrences of the query terms in the documents; (d) retrieving a subsequent portion of the inverted index corresponding to a subsequent plurality of documents from the first memory into the second memory; (e) searching the subsequent portion of the inverted index for each of the query terms and determining a number of occurrences of the query terms in each of the documents in the subsequent portion of the inverted index; (f) calculating similarity scores for each of the documents in the subsequent portion of the inverted index based on the number of occurrences of the query terms in the documents; and (g) repeating steps (d) through (f) until a similarity score has been calculated for each of the documents in the corpus of documents.

Claim 17

A computer implemented method for performing a search of a corpus of documents for similarity to a query having multiple terms using an inverted index of the corpus of documents, the method comprising the sequential steps of: (a) retrieving a first portion of the inverted index corresponding to a first plurality of documents from a first memory into a second memory; (b) searching the first portion of the inverted index for each of the query terms and determining a number of occurrences of the query terms in each of the documents in the first portion of the inverted index; (c) calculating similarity scores for each of the documents in the first portion of the inverted index based on the number of occurrences of the query terms in the documents; (d) retrieving a subsequent portion of the inverted index corresponding to a subsequent plurality of documents from the first memory into the second memory; (e) searching the subsequent portion of the inverted index for each of the query terms and determining a number of occurrences of the query terms in each of the documents in the subsequent portion of the inverted index; (f) calculating similarity scores for each of the documents in the subsequent portion of the inverted index based on the number of occurrences of the query terms in the documents; and (g) repeating steps (d) through (f) until a similarity score has been calculated for each of the documents in the corpus of documents.

Claim 18

The method of claim 14, further comprising the step of ranking the plurality of documents based on the calculated similarity scores.

Claim 19

The method of claim 14, further comprising the step of producing a list of the plurality documents that is ordered based on the calculated similarity scores.

Claim 20

The method of claim 14, wherein the query terms comprise images, and wherein the inverted index comprises an inverted index of images that occur in the plurality of documents.

Claim 21

The method of claim 14, further comprising the step of storing a predetermined number of the highest similarity scores in the second memory as the searching and calculating steps are performed.

Claim 22

The method of claim 14, wherein similarity scores for documents in each portion of the inverted index are calculated before a search is performed for occurrences of the query terms in documents in a subsequent portion of the inverted index.

Claim 23

The method of claim 14, wherein, the steps of calculating a similarity score comprise calculating a similarity score based on a user defined formula and a number of occurrences of the query terms in a document.