In this short note we present a recommendation system for automatic retrieval of broken Web links using an
approach based on contextual information. We extract information from the context of a link such as the anchor
text, the content of the page containing the link, and a combination of the cache page in some search engine and web
archive, if it exists. Then the selected information is processed and submitted to a search engine. We propose an
algorithm based on information retrieval techniques to select the most relevant information and to rank the candidate
pages provided for the search engine, in order to help the user to find the best replacement. To test the different
methods, we have also defined a methodology which does not require the user judgements, what increases the objectivity
of the results.