Predictive approach to URL determination
Embodiments of the present invention provide methods, computer program products, and systems for suggesting uniform resource locators located at previously unknown hosts to users. Embodiments of the present invention can be used to record uniform resource locators (URLs) associated with web resources accessed by a user and usage details pertaining to the user's usage of the web resources. Responsive to receiving a request to access a hostname not stored in the recorded URLs, embodiments of the present invention can be used to receive URLs associated with web resources available at the hostname and assign weight values to the received URLs based, at least in part, on the usage details and a similarity between the received URLs and the recorded URLs. Embodiments of the present invention can then be used to generate a URL suggestion from the received URLs based, at least in part, on the assigned weight values.
1. A computer-implemented method comprising:
recording uniform resource locators (URLs) associated with web resources accessed by a user and usage details pertaining to the user's usage of the web resources, wherein the usage details comprise browsing history of a user, an amount of time spent accessing one or more URLs, frequency with which the one or more URLs are accessed, and duration between accessing the one or more URLs;
responsive to receiving a request to access a hostname not stored in the recorded URLs from a partial URL typed by the user into an address bar of a web browser, receiving URLs associated with web resources available at the hostname;
assigning weight values to the received URLs based, at least in part, on the usage details and a similarity between the received URLs and the recorded URLs, wherein the similarity between the received URLs and the recorded URLs is based on a number of shared characters; and
generating a URL suggestion from the received URLs based, at least in part, on the assigned weight values, wherein the generated URL suggestion is displayed to the user in the address bar of the web browser;
wherein assigning weight values to the received URLs based, at least in part, on the usage details and a similarity between the received URLs and the recorded URLs comprises: ranking the received URLs according to their respective similarities to the recorded URLs, determining a subset of the received URLs based on their respective rankings, ranking the received URLs in the determined subset based on the usage details of their respectively similar recorded URLs, and assigning weight values to the received URLs in the determined subset based on their respective rankings in the subset.