URL structure : Hyphens or underscores URLs revisited

• Traditionally it isn’t seen by search engines as a word separator (this is slowly changing now)

• Hindari penggunaan pemisah kata underscores (_) pada subdomain dan sebagai gantinya gunakan tanda hypen (-).
• So if you have a url like word1_word2, Google will only return that page if the user searches for word1_word2 (which almost never happens). If you have a url like word1-word2, that page can be returned for the searches word1, word2, and even “word1 word2″.

Common causes of Underscores Problem

• Additive filtering of a set of items Many sites provide different views of the same set of items or search results, often allowing the user to filter this set using defined criteria (for example: show me hotels on the beach). When filters can be combined in a additive manner (for example: hotels on the beach and with a fitness center), the number of URLs (views of data) in the sites explodes. Creating a large number of slightly different lists of hotels is redundant, because Googlebot needs to see only a small number of lists from which it can reach the page for each hotel. For example:
  • Hotel properties at "value rates": http://www.example.com/hotel-search-results.jsp?Ne=292&N=461
  • Hotel properties at "value rates" on the beach: http://www.example.com/hotel-search-results.jsp?Ne=292&N=461+4294967240
  • Hotel properties at "value rates" on the beach and with a fitness center: http://www.example.com/hotel-search-results.jsp?Ne=292&N=461+4294967240+4294967270
• Dynamic generation of documents. This can result in small changes because of counters, timestamps, or advertisements.
• Problematic parameters in the URL. Session IDs, for example, can create massive amounts of duplication and a greater number of URLs.
• Sorting parameters. Some large shopping sites provide multiple ways to sort the same items, resulting in a much greater number of URLs. For example: http://www.example.com/results?search_type=search_videos&search_query=tpb&search_sort=relevance &search_category=25
• Irrelevant parameters in the URL, such as referral parameters. For example: http://www.example.com/search/noheaders?click=6EE2BF1AF6A3D705D5561B7C3564D9C2&clickPage= OPD+Product+Page&cat=79
  http://www.example.com/discuss/showthread.php?referrerid=249406&threadid=535913
  http://www.example.com/products/products.asp?N=200063&Ne=500955&ref=foo%2Cbar&Cn=Accessories.
• Calendar issues. A dynamically generated calendar might generate links to future and previous dates with no restrictions on start of end dates. For example: http://www.example.com/calendar.php?d=13&m=8&y=2011
  http://www.example.com/calendar/cgi?2008&month=jan
• Broken relative links. Broken relative links can often cause infinite spaces. Frequently, this problem arises because of repeated path elements. For example: http://www.example.com/index.shtml/discuss/category/school/061121/html/interview/ category/health/070223/html/category/business/070302/html/category/community/070413/html/FAQ.htm

• Consider using a robots.txt file to block Googlebot's access to problematic URLs. Typically, you should consider blocking dynamic URLs, such as URLs that generate search results, or URLs that can create infinite spaces, such as calendars. Using regular expressions in your robots.txt file can allow you to easily block large numbers of URLs.
• Wherever possible, avoid the use of session IDs in URLs. Consider using cookies instead. Check our Webmaster Guidelines for additional information.
• Whenever possible, shorten URLs by trimming unnecessary parameters.
• If your site has an infinite calendar, add a nofollow attribute to links to dynamically created future calendar pages.
• Check your site for broken relative links.