Reading difficulty level based resource recommendation

A system including a user interface for a search engine, the user interface of a computer comprising:

• input controls with which a user designates content of interest within a source document;
• a document acquisition module to obtain candidate resources from a search engine based on the user-designated content of the source document;
• a reading difficulty level module to generate reading difficulty scores for each of the candidate resources;
• a subject matter similarity module to generate similarity scores between each of the candidate resources and the source document;
• and
• a recommendation module to recommend a subset of the candidate resources based on both of the reading difficulty scores and the similarity scores, the user interface displaying the subset of candidate resources with display of the source document;
• wherein the reading difficulty level module is to generate a specialized reading difficulty score based on the candidate resource being specific to a particular field, the specialized reading difficulty score based on standards in that particular field;
• and
• wherein the subject matter similarity module is to generate a feature vector for each of the source document and candidate resources, the feature vectors being compared to determine subject matter similarity.

A method, comprising:

• extracting content from a source document in response to a user interaction with the source document;
• obtaining, using the content, candidate resources related to the source document;
• evaluating the reading difficulty level of the candidate resources based on content extracted from the candidate resources, including calculating a probability distribution of words across a number of different subject matters when generating the reading difficulty level;
• presenting, to the user, a recommended candidate resource, where the recommended candidate resource is selected based on the reading difficulty level of the candidate resources, a reading difficulty level of the source document, the content of the source document, and content of the candidate resources.