Granted Patent B1

US 7,739,253 · App. 11/112,505 · Granted Jun 15, 2010

Abstract

Claims (104)

1. A computer implemented method, comprising:

determining statistics for each of a plurality of Web pages with respect to one or more predetermined categories based on content rating of each of the Web pages; and

for each of the predetermined categories, selecting a set of primary Web pages having relationships with one or more secondary Web pages, wherein the one or more secondary Web pages probabilistically distinguish from relationships with a remainder of the secondary Web pages, wherein the selected set of primary Web pages having relationships with one or more secondary Web pages are used by a network access device to determine a probability that a particular Web page received at the network access device responsive to a request by a client belongs to one or more of the predetermined categories based on a linking pattern of the particular Web page linked with other Web pages using content ratings of the selected set of primary Web pages and the associated secondary Web pages,

wherein the network access device is a gateway device that interfaces the client of a local area network (LAN) with a Web server of an external network, wherein the particular Web page is routed exclusively through the gateway device from the external network to the client with the gateway device acting as a sole interface between the LAN and the external network for the purpose of downloading a Web site, and wherein the determined probability of the particular Web page is used by the network access device to determine whether the particular Web page should be delivered to the client over the LAN.

2. The method of claim 1 , further comprising assigning a probability to each of the primary Web pages having relationships with one or more secondary Web pages to one or more of the predetermined categories to represent a likelihood of a Web page having similar configuration is associated with a respective category, wherein the probability is determined based on a number of occurrence of a linked-to category in training data associated with each category and in a particular Web page being categorized.

3. The method of claim 2 , wherein the probability is determined based on a summation of probabilities calculated using following formula:

∑

s

∈

S

⁢

P

i

⁡

(

W

❘

T

,

s

)

=

P

i

⁡

(

W

❘

s

)

⁢

P

i

⁡

(

T

❘

W

,

s

)

P

i

⁡

(

T

❘

s

)

wherein P i (X, s) is the probability at which a linked-to category number (s) occurs in X which is a set of input Web pages for category (i), wherein the linked-to category number (s) is greater than one, wherein T is a set of training data and W is a Web page being categorized, and wherein S is a set of all linked-to category number (s) that appear in the training data (T) or the Web page (W).

4. The method of claim 3 , further comprising:

storing in a database the selected primary and secondary Web pages as a trained set of Web pages for rating a subsequent Web page; and

storing in the database relationships between the primary and secondary Web pages as one or more rules, wherein the one or more rules are used to categorize the subsequent Web page having similar relationships with one or more selected secondary Web pages into one or more of the predetermined categories.

5. The method of claim 4 , further comprising:

receiving at the network access device a first Web page from a Web server over the external network, the first Web page being requested from a client over the LAN, the first Web page having one or more links linking with one or more second Web pages;

the network access device accessing the one or more links of the first Web page to obtain the one or more second Web pages;

the network access device accessing the database to obtain content rating for each of the second Web pages based on a trained set of Web pages; and

the network access device determining a final rating for the first Web page based on the content rating of each of the second Web pages.

6. The method of claim 5 , further comprising the network access device accessing the database to obtain content rating of the first Web page based on the content of the first Web page, where the final rating of the first Web page is determined further based on the content rating of the first Web page.

7. The method of claim 5 , further comprising the network access device determining whether the client is allowed to access the first Web page based on the final rating of the first Web page.

8. The method of claim 7 , further comprising:

if the database does not contain content rating of any one of the first and second Web pages, performing content rating on any one of unrated first and second Web pages; and

storing the content ratings of the unrated Web pages in the database as a part of trained set of Web pages.

9. The method of claim 1 , wherein a relationship between a primary Web page of the plurality of primary Web pages and a secondary Web page of the one or more secondary Web pages is determined based on a link linking with the primary and secondary Web pages and content of the primary and secondary Web pages.

10. The method of claim 9 , wherein the link is a URL (universal resource locator) compatible link within the primary Web page linking with the secondary Web page, and wherein the external network is part of the Internet.

11. The method of claim 9 , further comprising:

performing content rating on the primary Web page; and

for each of the linked secondary Web pages of the primary Web page, accessing the respective secondary Web page via the respective link and performing content rating on the respective secondary Web page.

12. The method of claim 11 , further comprising determining a probability of the primary Web page linking with the secondary Web pages based on the contenting rating of the primary and secondary Web pages, wherein the probability of the primary Web page linking with the secondary Web pages represents a likelihood that the primary Web page belongs to one of the predetermined categories.

13. The method of claim 1 , wherein the gateway device comprises a modem.

14. A machine-readable storage medium having executable code stored thereon to cause a machine to perform a method, the method comprising:

determining statistics for each of a plurality of Web pages with respect to one or more predetermined categories based on content rating of each of the Web pages; and

for each of the predetermined categories, selecting a set of primary Web pages having relationships with one or more secondary Web pages, wherein the one or more secondary Web pages probabilistically distinguish from relationships with a remainder of the Web pages, wherein the selected set of primary Web pages having relationships with one or more secondary Web pages are used by a network access device to determine a probability that a particular Web page received at the network access device responsive to a request by a client belongs to one or more of the predetermined categories based on a linking pattern of the particular Web page linked with other Web pages using content ratings of the selected set of primary Web pages and the associated secondary Web pages,

wherein the network access device is a gateway device that interfaces the client of a local area network (LAN) with a Web server of an external network, wherein the particular Web page is routed exclusively through the gateway device from the external network to the client with the gateway device acting as a sole interface between the LAN and the external network for the purpose of downloading a Web site and wherein the determined probability of the particular Web page is used by the network access device to determine whether the particular Web page should be delivered to the client over the LAN.

15. The machine-readable storage medium of claim 14 , wherein the gateway device comprises a modem.

16. A system, comprising:

a processor; and

a memory coupled to the processor to store instructions, when executed from the memory, cause the processor to perform operations including

determining statistics for each of a plurality of Web pages with respect to one or more predetermined categories based on content rating of each of the Web pages, and

for each of the predetermined categories, selecting a set of primary Web pages having relationships with one or more secondary Web pages, wherein the one or more secondary Web pages probabilistically distinguish from relationships with a remainder of the Web pages, wherein the selected set of primary Web pages having relationships with one or more secondary Web pages are used by a network access device to determine a probability that a particular Web page received at the network access device responsive to a request by a client belongs to one or more of the predetermined categories based on a linking pattern of the particular Web page linked with other Web pages using content ratings of the selected set of primary Web pages and the associated secondary Web pages,

wherein the network access device is a gateway device that interfaces the client of a local area network (LAN) with a Web server of an external network, wherein the particular Web page is routed exclusively through the gateway device from the external network to the client with the gateway device acting as a sole interface between the LAN and the external network for the purpose of downloading a Web site, and wherein the determined probability of the particular Web page is used by the network access device to determine whether the particular Web page should be delivered to the client over the LAN.

17. The system of claim 16 , wherein the gateway device comprises a modem.

18. A computer implemented method, comprising:

in response to a first Web page requested from a client and the first Web page having one or more links linking with one or more second Web pages, a network access device receiving the first Web page responsive to a request by the client and accessing the one or more links of the first Web page to obtain the one or more second Web pages, wherein the network access device is a gateway device interfacing the client over a local area network (LAN) with an external network, and wherein the first Web page and the one or more second Web pages are routed exclusively through the gateway device from the external network to the client with the gateway device acting as a sole interface between the LAN and the external network for the purpose of downloading a Web site;

the network access device accessing a database to obtain content rating for each of the second Web pages based on a trained set of Web pages; and

determining a final rating for the first Web page based on the content rating of each of the second Web pages, wherein the final rating of the first Web page indicates a probability that the first Web page belongs to a predetermined category based on a linking pattern of the first Web page with the second Web pages using the content rating of the second Web pages in view of the trained set of the Web pages, wherein the determined final rating associated with the first Web page is used by the network access device to determine whether the first Web page should be delivered to the client over the LAN.

19. The method of claim 18 , further comprising accessing the database to obtain content rating of the first Web page based on the content of the first Web page, where the final rating of the first Web page is determined further based on the content rating of the first Web page.

20. The method of claim 18 , further comprising the network access device determining whether the client is allowed to access the first Web page based on the final rating of the first Web page.

21. The method of claim 20 , further comprising:

if the database does not contain content rating of any one of the first and second Web pages, performing content rating on any one of unrated first and second Web pages; and

storing the content ratings of the unrated Web pages in the database as a part of trained set of Web pages.

22. The method of claim 18 , wherein the gateway device comprises a modem.

23. A machine-readable storage medium having executable code stored thereon to cause a machine to perform a method, the method comprising:

in response to a first Web page requested from a client and the first Web page having one or more links linking with one or more second Web pages, a network access device receiving the first Web page responsive to a request by the client and accessing the one or more links of the first Web page to obtain the one or more second Web pages, wherein the network access device is a gateway device interfacing the client over a local area network (LAN) with an external network, and wherein the first Web page and the one or more second Web pages are routed exclusively through the gateway device from the external network to the client with the gateway device acting as a sole interface between the LAN and the external network for the purpose of downloading a Web site;

accessing a database to obtain content rating for each of the second Web pages based on a trained set of Web pages; and

determining a final rating for the first Web page based on the content rating of each of the second Web pages, wherein the final rating of the first Web page indicates a probability that the first Web page belongs to a predetermined category based on a linking pattern of the first Web page with the second Web pages using the content rating of the second Web pages in view of the trained set of the Web pages, wherein the determined final rating associated with the first Web page is used by the network access device to determine whether the first Web page should be delivered to the client over the LAN.

24. The machine-readable storage medium of claim 23 , wherein the gateway device comprises a modem.

25. A system, comprising:

a processor; and

a memory coupled to the processor to store instructions, when executed from the memory, cause the processor to perform operations including

in response to a first Web page requested from a client and the first Web page having one or more links linking with one or more second Web pages, a network access device receiving the first Web page responsive to a request by the client and accessing the one or more links of the first Web page to obtain the one or more second Web pages, wherein the network access device is a gateway device interfacing the client over a local area network (LAN) with an external network, and wherein the first Web page and the one or more second Web pages are routed exclusively through the gateway device from the external network to the client with the gateway device acting as a sole interface between the LAN and the external network for the purpose of downloading a Web site,

accessing a database to obtain content rating for each of the second Web pages based on a trained set of Web pages, and

determining a final rating for the first Web page based on the content rating of each of the second Web pages, wherein the final rating of the first Web page indicates a probability that the first Web page belongs to a predetermined category based on a linking pattern of the first Web page with the second Web pages using the content rating of the second Web pages in view of the trained set of the Web pages, wherein the determined final rating associated with the first Web page is used by the network access device to determine whether the first Web page should be delivered to the client over the LAN.

26. The system of claim 25 , wherein the gateway device comprises a modem.

Assignments (23)