System and method for incorporating anchor text into ranking search results

• US 20060074871A1
• Filed: 09/30/2004
• Published: 04/06/2006
• Est. Priority Date: 09/30/2004
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Meyerzon, Dmitriy, Taylor, Michael J., Zaragoza, Hugo, Robertson, Stephen Edward
• Granted Patent: US 7,739,277 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9538 Presentation of query results

Abstract

Search results of a search query on a network are ranked according to a scoring function that incorporates anchor text as a term. The scoring function is adjusted so that a target document of anchor text reflect the use of terms in the anchor text in the target document'"'"'s ranking. Initially, the properties associated with the anchor text are collected during a crawl of the network. A separate index is generated that includes an inverted list of the documents and the terms in the anchor text. The index is then consulted in response to a query to calculate a document'"'"'s score. The score is then used to rank the documents and produce the query results.

Claims

1. A computer-implemented method for ranking search results, comprising:

recording anchor text properties from a network, wherein the anchor text properties correspond to a source document and a target document on the network;

generating an anchor text index that associates each target document with a term included in the anchor text, wherein the anchor text index is separate from a content index;

accessing the anchor text index to determine quantities associated with the anchor text of a particular target document in the index; and

using the anchor text quantities to adjust a scoring function that is used to rank the documents for producing the search results.

2. The computer-implemented method of claim 1, wherein recording anchor text properties further comprises generating a representation of the network from recorded document and link information.

3. The computer-implemented method of claim 2, further comprising using the representation of the network to populate a table with the anchor text properties.

4. The computer-implemented method of claim 1, wherein the anchor text properties include at least one of a source identifier, a target identifier, anchor text, and a link identifier.

5. The computer-implemented method of claim 1, further comprising ranking the documents according to a scoring function (score) that is determined according to at least:

a weighted anchor text term frequency (wtf_Anchor) and an anchor text length normalization component (B_Anchor).

6. The computer-implemented method of claim 5, wherein the anchor text length normalization component (B_Anchor) corresponds to an anchor text weighted document length (wdl), an average weighted document length (avwdl).

7. The computer-implemented method of claim 1, further comprising ranking the documents according to a scoring function (score) that is determined according to at least:

a weighted term frequency (wtf), a weighted anchor text term frequency (wtf_Anchor), a length normalization component, an anchor text length normalization component (B_Anchor) an average weighted document length (avwdl), a number of documents on the network (N);

a number of documents containing a query term (n), and a constant (k₁).

8. The computer-implemented method of claim 7, wherein the scoring function (score) is given by:

$\mathrm{score}=\sum <br><br>\frac{\left(\frac{\mathrm{wtf}}{B}+\frac{{\mathrm{wtf}}_{\mathrm{Anchor}}}{B_{\mathrm{Anchor}}}\right)<br><br>\left(k_1+1\right)}{k_1+\left(\frac{\mathrm{wtf}}{B}+\frac{{\mathrm{wtf}}_{\mathrm{Anchor}}}{B_{\mathrm{Anchor}}}\right)}\times <br><br>\log <br><br>\left(\frac{N}{n}\right)$

9. The computer implemented method of claim 8, wherein B_Anchor differs from B by taking components of B_Anchor from the anchor text index and taking components from B from the content index.

10. The computer implemented method of claim 8, wherein the strength of the length normalization provided by B_Anchor is adjusted by choosing a different constant value associated with B_Anchor.

11. A system for ranking search results, comprising:

a search engine included on a computing device, the search engine configured to execute computer-executable instructions, the computer-executable instructions comprising;

crawling a network to discover documents on the network;

recording anchor text properties from a network, wherein the anchor text properties correspond to a source document and a target document on the network;

generating an anchor text index that associates each target document with a term included in the anchor text, wherein the anchor text index is separate from a content index;

accessing the anchor text index to determine quantities associated with the anchor text of a particular target document in the index; and

using the anchor text quantities to adjust a scoring function that is used to rank the documents for producing the search results.

12. The system of claim 11, wherein recording anchor text properties further comprises generating a representation of the network from recorded document and link information.

13. The system of claim 12, further comprising using the representation of the network to populate a table with the anchor text properties.

14. The system of claim 11, further comprising ranking the documents according to a scoring function (score) that is determined according to at least:

a weighted anchor text term frequency (wtf_Anchor) and an anchor text length normalization component (B_Anchor)

15. The system of claim 11, further comprising ranking the documents according to a scoring function (score) that is determined according to at least:

a weighted term frequency (wtf), a weighted anchor text term frequency (wtf_Anchor), a length normalization component, an anchor text length normalization component (B_Anchor) an average weighted document length (avwdl), a number of documents on the network (N);

a number of documents containing a query term (n), and a constant (k₁).

16. The system of claim 15, wherein the scoring function (score) is given by:

$\mathrm{score}=\sum <br><br>\frac{\left(\frac{\mathrm{wtf}}{B}+\frac{{\mathrm{wtf}}_{\mathrm{Anchor}}}{B_{\mathrm{Anchor}}}\right)<br><br>\left(k_1+1\right)}{k_1+\left(\frac{\mathrm{wtf}}{B}+\frac{{\mathrm{wtf}}_{\mathrm{Anchor}}}{B_{\mathrm{Anchor}}}\right)}\times <br><br>\log <br><br>\left(\frac{N}{n}\right)$

17. The system of claim 16, wherein B_Anchor differs from B by taking components of B_Anchor from the anchor text index and taking components from B from the content index.

18. The system of claim 16, wherein the strength of the length normalization provided by B_Anchor is adjusted by choosing a different constant value associated with B_Anchor.

19. A computer-readable medium that includes computer-executable instructions for ranking search results, the instructions comprising:

storing document and link information for documents on a network;

generating a representation of the network from the recorded document and link information;

recording anchor text properties from the representation of the network;

wherein the anchor text properties are used to populate a table that includes a reference to a source document and a target document associated with the anchor text;

generating an anchor text index that associates each target document with a term included in the anchor text, wherein the anchor text index is separate from a content index;

accessing the anchor text index to determine quantities associated with the anchor text of a particular target document in the index; and

using the anchor text quantities to adjust a scoring function that is used to rank the documents for producing the search results.

20. The computer-readable medium of claim 19, wherein the anchor text properties include at least one of a source identifier, a target identifier, anchor text, and a link identifier.

21. The computer-readable medium of claim 19, further comprising ranking the documents according to a scoring function (score) that is determined according to at least:

a weighted anchor text term frequency (wtf_Anchor) and an anchor text length normalization component (B_Anchor).

22. The computer-readable medium of claim 21, wherein the anchor text length normalization component (B_Anchor) corresponds to an anchor text weighted document length (wdl), an average weighted document length (avwdl).

23. The computer-readable medium of claim 19, further comprising ranking the documents according to a scoring function (score) that is determined according to at least:

a weighted term frequency (wtf), a weighted anchor text term frequency (wtf_Anchor), a length normalization component, an anchor text length normalization component (B_Anchor) an average weighted document length (avwdl), a number of documents on the network (N);

a number of documents containing a query term (n), and a constant (k₁).

24. The computer-readable medium of claim 23, wherein the scoring function (score) is given by:

$\mathrm{score}=\sum <br><br>\frac{\left(\frac{\mathrm{wtf}}{B}+\frac{{\mathrm{wtf}}_{\mathrm{Anchor}}}{B_{\mathrm{Anchor}}}\right)<br><br>\left(k_1+1\right)}{k_1+\left(\frac{\mathrm{wtf}}{B}+\frac{{\mathrm{wtf}}_{\mathrm{Anchor}}}{B_{\mathrm{Anchor}}}\right)}\times <br><br>\log <br><br>\left(\frac{N}{n}\right)$

25. The computer-readable medium of claim 24, wherein B_Anchor differs from B by taking components of B_Anchor from the anchor text index and taking components from B from the content index.

26. The computer-readable medium of claim 24, wherein the strength of the length normalization provided by B_Anchor is adjusted by choosing a different constant value associated with B_Anchor.