Controlling access to a website

1. A method of website access control, comprising:

accessing, by a communications unit of a user terminal, a plurality of training websites over a network;

classifying, by a classification unit of the user terminal, the training websites in a content category based on a user input;

extracting, by a feature extraction unit, one or more features indicative of the content category from the training websites; determining, by a classifier determining unit, a classifier for classifying websites in the content category based on the extracted features;

verifying, by a verification unit, the determined classifier by accessing, by the communications unit of the user terminal, a plurality of verification websites and classifying the verification websites using the classifier and if the verification websites are classified in the content category, prompting a user to verify the classification of the verification websites; and

classifying, by an access control unit, further requested websites using the verified determined classifier and controlling access to the further requested websites based on the classification of the websites, wherein the feature extraction unit, the classifier determining unit and the verification unit form part of the user terminal, the method further comprising:

transmitting the verified determined classifier to a server and storing the classifier and a corresponding set of training websites in a memory of the server as part of a plurality of classifiers and a plurality of associated sets of training websites;

comparing, by a classifier comparison unit, the plurality of classifiers to determine whether the content categories of each overlap;

if the content categories of the plurality of classifiers overlap, extracting, by the feature extraction unit, a plurality of training websites from each of the sets of training websites corresponding to the overlapping classifiers;

merging, by the classifier determining unit, the extracted plurality of training websites into a global set of training websites; and

determining, by the classifier determining unit, a global classifier based on the global set of training websites.

2. A method according to claim 1 , wherein controlling access to the further requested websites comprises denying access if the classifier classifies the website in the content category.

3. A method according to claim 1 , further comprising,

if the user does not verify the classification of the verification websites, extracting features indicative of the content category from the verification website and updating the classifier based on the extracted features.

4. A method according to claim 3 , wherein the classifier comprises an initial classifier and a control classifier, the initial classifier being determined based on the features extracted from the training websites, and the control classifier being determined based on the features extracted from the verification websites, wherein updating the classifier comprises determining the control classifier based on the features extracted from the verification websites.

5. A method according to claim 4 , wherein, the verification websites are classified using the initial classifier and, if the initial classifier classifies the verification website in the content category, the classification is checked using the control classifier, the user being prompted if the initial classifier and the control classifier classify the verification website in the content category.

6. A method according to claim 4 , wherein controlling access to a website comprises:

classifying the website using the initial classifier;

if the initial classifier classifies the further website in the content category, checking the classification using the control classifier; and controlling access to the website if the initial classifier and the control classifier classify the website in the content category.

7. A method according to claim 1 , further comprising, if the user verifies the classification, incrementing a correct counter and, if the user does not verify the classification, incrementing an incorrect counter.

8. A method according to claim 7 , further comprising verifying the classifier if the ratio of correct classifications to incorrect classifications reaches a threshold value.

9. A method according to claim 1 , further comprising at least one of repeating the verification at a predetermined time; and repeating the verification based on a user input.

10. A method according to claim 1 , wherein determining the classifier comprises selecting features present in a number of the plurality of training websites greater than or equal to a valid feature threshold and including the selected features in the classifier.

11. A method according to claim 1 , wherein the feature extraction unit, the classifier determining unit and the verification unit form part of the server, the method further comprising storing the classifier and a corresponding set of training websites in a memory of the server.

12. A method according to claim 1 , further comprising transmitting, by the communications unit a notification generated by the classifier determining unit recommending the global classifier to one or more user terminals using one or more of the plurality of classifiers.

13. A method according to claim 1 , wherein the access control unit forms part of the server, the method further comprising:

comparing, by a classifier comparison unit, a further website classified based on the determined classifier with a plurality of training websites corresponding to an undetermined classifier;

if the further website matches one of the plurality of training websites for the undetermined classifier, retrieving, by the classifier comparison unit, one or more training websites corresponding to the determined classifier;

merging, by the classifier determining unit, the retrieved one or more training websites corresponding to the determined classifier and the training websites corresponding to the undetermined classifier into a global set of training websites; and

determining, by the classifier determining unit, a global classifier based on the global set of training websites.

14. A method according to claim 1 , wherein the access control unit forms part of the server, the method further comprising, at the server:

receiving, by the communications unit, data from a further user terminal relating to one or more training websites corresponding to an undetermined classifier;

classifying, by the access control unit, the one or more training websites corresponding to the undetermined classifier based on the determined classifier;

if the one or more training websites corresponding to the undetermined classifier is classified in the content category of the determined classifier, transmitting, by the communications unit a notification generated by the classifier determining unit recommending the global classifier to the further user terminal.

15. A non-transitory computer readable medium comprising computer readable code configured to carry out a method of website access control, the method comprising:

accessing, by a communications unit of a user terminal, a plurality of training websites over a network; classifying, by a classification unit of the user terminal, the training websites in a content category based on a user input; extracting, by a feature extraction unit, one or more features indicative of the content category from the training websites;

determining, by a classifier determining unit, a classifier for classifying websites in the content category based on the extracted features;

verifying, by a verification unit, the determined classifier by accessing, by the communications unit of the user terminal, a plurality of verification websites and classifying the verification websites using the classifier and if the verification websites are classified in the content category, prompting a user to verify the classification of the verification websites; and

classifying, by an access control unit, further requested websites using the verified determined classifier and controlling access to the further requested websites based on the classification of the websites, wherein the feature extraction unit, the classifier determining unit and the verification unit form part of the user terminal, the method further comprising:

transmitting the verified determined classifier to a server and storing the classifier and a corresponding set of training websites in a memory of the server as part of a plurality of classifiers and a plurality of associated sets of training websites;

comparing, by a classifier comparison unit, the plurality of classifiers to determine whether the content categories of each overlap;

if the content categories of the plurality of classifiers overlap, extracting, by the feature extraction unit, a plurality of training websites from each of the sets of training websites corresponding to the overlapping classifiers;

merging, by the classifier determining unit, the extracted plurality of training websites into a global set of training websites; and

determining, by the classifier determining unit, a global classifier based on the global set of training websites.

16. A user terminal for controlling access to a website, the user terminal comprising:

a communication unit configured to access a plurality of training websites over the network; a classification unit configured to train the user terminal by classifying the training websites in a content category based on a user input;

a feature extraction unit configured to extract from the training websites one or more features indicative of the content category; a classifier determining unit configured to determine a classifier for classifying websites in the content category based on the extracted features; wherein the communication unit is further configured to access a plurality of verification websites and a verification unit is configured to verify the determined classifier by classifying the verification websites using the classifier and if the verification websites are classified in the content category, prompting a user to verify the classification of the verification websites; and

an access control unit configured to classify further requested websites using the verified determined classifier and to control access to the further requested websites based on a classification of the websites, wherein the feature extraction unit, the classifier determining unit and the verification unit form part of the user terminal, further comprising:

transmitting the verified determined classifier to a server and storing the classifier and a corresponding set of training websites in a memory of the server as part of a plurality of classifiers and a plurality of associated sets of training websites;

comparing, by a classifier comparison unit, the plurality of classifiers to determine whether the content categories of each overlap;

if the content categories of the plurality of classifiers overlap, extracting, by the feature extraction unit, a plurality of training websites from each of the sets of training websites corresponding to the overlapping classifiers;

merging, by the classifier determining unit, the extracted plurality of training websites into a global set of training websites; and

determining, by the classifier determining unit, a global classifier based on the global set of training websites.

17. A server for controlling access to a website, the server comprising:

a communication unit configured to receive data relating to a plurality of training websites over the network;

a feature extraction unit configured to extract from the training websites one or more features indicative of the content category;

a classifier determining unit configured to determine a classifier for classifying websites in the content category based on the extracted features;

wherein the communication unit is further configured to receive data relating to a plurality of verification websites and a verification unit is configured to verify the determined classifier by classifying the verification websites using the classifier and if the verification websites are classified in the content category, prompting a user of a user equipment to verify the classification of the verification websites; and an access control unit configured to classify further requested websites using the verified determined classifier and to control access to the further requested websites based on a classification of the websites, wherein the feature extraction unit, the classifier determining unit and the verification unit form part of the server, further comprising:

storing the classifier and a corresponding set of training websites in a memory of the server as part of a plurality of classifiers and a plurality of associated sets of training websites;

comparing, by a classifier comparison unit, the plurality of classifiers to determine whether the content categories of each overlap;

if the content categories of the plurality of classifiers overlap, extracting, by the feature extraction unit, a plurality of training websites from each of the sets of training websites corresponding to the overlapping classifiers;

merging, by the classifier determining unit, the extracted plurality of training websites into a global set of training websites; and

determining, by the classifier determining unit, a global classifier based on the global set of training websites.