System and method for measuring semantic relevance of keywords

摘要

The server may receive the seed keyword and generate additional keywords associated with the seed keyword. The server can use the semantic relationship graph to identify keyword categories. Each keyword may have a semantic distance to the seed keyword less than the threshold. The server can generate a keyword-seed affinity score for each keyword of the keyword category based on the frequency of the keyword generated with the seed keyword on the information resource. The server may determine a category-seed affinity score for each keyword category based on a keyword-seed affinity score for each of the keywords in the keyword category. The server can compare each category-seed affinity score with a threshold value. The server can send keywords for display. Based on the comparison, one keyword category may be marked as selected and another keyword category as non-selected.