Identifying a document by performing spectral analysis on the contents of the document
A system and method for identifying a document based on a spectral analysis of the text of the document is described. In some examples, the system generates a document identifier for a rendered document based on assigning values to words in the rendered document, such as values associated with the frequency of use of the word by the rendered document, the absolute or relative position of the word in the rendered document, and so on. The system may use the document identifier to generate a group of documents having similar document identifiers, and choose a likely match from the group of documents based on predictive analysis.
1. A method, comprising:
creating an index, with one or more entries, for a rendered document at a computer system, wherein each of the one or more index entries contains a word and one or more values, wherein each value represents a characteristic of the word;
selecting a subset of the one or more index entries from the index using the computer system;
creating an ordered sequence of values from the selected subset of the one or more index entries, wherein each value comprises a position of the entry relative to another entry in the selected subset of the one or more index entries;
building an identifier for the rendered document at the computer system based on the ordered sequence of the values, wherein each index entry in the subset satisfies an identifier rule for a document identifier; and
locating an electronic counterpart to the rendered document based on the identifier for the rendered document using the computer system.
2. The method of claim 1 , wherein a value of an index entry is associated with a frequency of instances of the word contained with the index entry within the rendered document.
3. The method of claim 1 , wherein a value of an index entry is associated with an absolute position of the word contained with the index entry within the rendered document.
4. The method of claim 1 , wherein a value of an index entry is associated with a relative position of the word contained with the index entry within the rendered document.
5. The method of claim 1 , further comprising:
performing a hash function on the identifier for the rendered document to generate a hash value for the rendered document; and
designating the hash value as the identifier of the rendered document.
6. The method of claim 1 , further comprising:
querying a plurality of documents to identify one or more documents, each of the one or more documents associated with an identifier that at least partially matches the identifier for the rendered document.
7. The method of claim 1 , wherein the ordered sequence of the values from the selected subset of the one or more index entries in the selected subset comprises a sequence of the values of the one or more index entries in the selected subset that is ordered based on absolute positions in the rendered document of the words contained with the one or more index entries in the selected subset.
8. The method of claim 1 , wherein the ordered sequence of the values from the selected subset of the one or more index entries in the selected subset comprises a sequence of the values of the one or more index entries in the selected subset that is ordered based on positional offsets in the rendered document of the words contained with the index entries in the selected subset.
9. The method of claim 1 , wherein the identifier rule comprises a rule to select index entries corresponding to words that appear N times or fewer within the rendered document, and wherein N >0.
10. The method of claim 1 , wherein building the identifier comprises determining offsets between the values of the ordered sequence of the values of the one or more index entries in the selected subset.
11. The method of claim 1 , wherein building the identifier comprises:
determining an ordinal sequence for the words in the ordered subset of the one or more index entries;
replacing each word in the ordinal sequence with a replacement value representing an initial appearance of the word within the rendered document; and
building the identifier using the replacement values.
12. An article of manufacture comprising a non-transitory computer-readable medium having instructions stored thereon that, if the instructions are executed by a computer, cause the computer to perform functions comprising:
creating an index, with one or more entries, for a rendered document, wherein each of the one or more index entries contains a word and one or more values, wherein each value represents a characteristic of the word;
selecting a subset of the one or more index entries from the index;
creating an ordered sequence of values from the selected subset of one or more index entries, wherein each value comprises a position of the entry relative to another entry in the selected subset of the one or more index entries;
building an identifier for the rendered document based on the ordered sequence of the values, wherein each index entry in the subset satisfies an identifier rule for a document identifier; and
locating an electronic counterpart to the rendered document based on the identifier for the rendered document.
13. The article of manufacture of claim 12 , wherein a value of an index entry is associated with a frequency of instances of the word contained with the index entry within the rendered document.
14. The article of manufacture of claim 12 , wherein a value of an index entry is associated with an absolute position of the word contained with the index entry within the rendered document.
15. The article of manufacture of claim 12 , wherein a value of an index entry is associated with a relative position of the word contained with the index entry within the rendered document.
16. The article of manufacture of claim 12 , wherein the functions further comprise:
performing a hash function on the identifier for the rendered document to generate a hash value for the rendered document; and
designating the hash value as the identifier of the rendered document.
17. The article of manufacture of claim 12 , wherein the functions further comprise:
querying a plurality of documents to identify one or more documents, each of the one or more documents associated with an identifier that at least partially matches the identifier for the rendered document.
18. The article of manufacture of claim 12 , wherein the ordered sequence of the values from the selected subset of the one or more index entries in the selected subset comprises a sequence of the values of the one or more index entries in the selected subset that is ordered based on absolute positions in the rendered document of the words contained with the one or more index entries in the subset.
19. The article of manufacture of claim 12 , wherein the ordered sequence of the values from the selected subset of the one or more index entries in the selected subset comprises a sequence of the values of the one or more index entries in the selected subset that is ordered based on positional offsets in the rendered document of the words contained with the index entries in the selected subset.