Recommended Methods for associating RSS

Http://web2.0focus.com /? P = 44

The associated RSS is one or more RSS related to a specific RSS content. In RSS subscription, obtaining the associated RSS not only saves the time for users to search for RSS, but also extends the range of RSS subscriptions for readers.

This method is similar to keyword-based search on search engines. After a keyword is input, the search engine also recommends related keywords while outputting the search results. When we look for tags on delicious, we will also give related tags. The preceding Association keywords or associated tags are usually relatively accurate. We can also use the associated method to make RSS recommendations.

In general, we need a powerful RSS database, preferably a closed system. For example, an RSS reading system has user data that can be used for statistics.

The foothold of our associated RSS is that the RSS subscribed by a user is generally concentrated in interest. Starting from an RSS feed, we obtain all users who have subscribed to this RSS feed and collect statistics on all users related to this RSS feed (more accurately, is all RSS of the RSS category, based on the type set by the user), the common description is "users who have subscribed to this RSS, and subscribed to this category ...".

For example, if the distribution of the RSS/RSS quantity calculated by the preceding method is Gaussian distribution, the RSS with a large statistical amount is related RSS. When users subscribe to RSS, they can list a certain number of RSS recommendations for users.

Another way to get associated RSS is that the RSS search engine processes its own data and obtains the associated RSS of a certain RSS. This requires the RSS search engine instead of blog search to collect massive RSS data.

The first step for search engines is to obtain RSS data and analyze RSS data. In general, RSS with an RSS entry description is sufficient to reflect the characteristics of RSS, and even RSS with only the RSS entry title is available. Analyze (Chinese RSS requires Word Segmentation) to obtain RSS-weighted keywords. In the above process, if there are enough resources, you can further obtain the content of the article referred to by RSS for further analysis.

The second step is the pre-processing process of RSS data, that is, through the Similarity Analysis of RSS keywords, obtain a normalized (normalized) RSS-RSS similarity value, and finally obtain the most similar rss list of a rss. This process
It also requires a considerable amount of resources, so using some RSS as seed RSS can save computing time. Fortunately, RSS Similarity analysis has less real-time requirements. After an associated RSS is analyzed, the "validity period" can last for a long time.

For a blog user, okrss uses the method of analyzing the blog user page to obtain the associated RSS. This method assumes that the blog user is closely related to the RSS mentioned by him. This method is easier for users who have links to RSS on their blogs to obtain associated RSS. However, for users who have only links to other users' blog URLs, they need to use the massive RSS database collected by okrss. This method uses RSS to compare the RSS-blog user homepage, collects and analyzes user logs, obtains possible URLs, and maps the URLs back to RSS to obtain the associated RSS of RSS.