Automatic keyword prediction using Google similarity distance

作者：Chen Ping I^*; Lin Shi Jen

来源：Expert Systems with Applications, 2010, 37(3): 1928-1938.

DOI：10.1016/j.eswa.2009.07.016

• Search engines
• Prediction
• Recommendation

In this paper, we present a new approach to help users using search engines without entering any keywords. What we want to do is to predict what word the users may want to search before they think about it. Most of the studies done in this field focus on how to help users enter keywords or how to re-rank the search results in order to make them more precise. Both of those methods need to establish a user behavior model and a repository in which to save the logs. In our proposed method, we use the Google similarity distance to measure keywords in the Webpage to find the potential keywords for the users. Thus, we do not need any repository. All the executions are on-line and real-time. Then, we extract all the important keywords as the potential search keywords. In this way, we can use these professional keywords to achieve precise search results. We believe that this can be useful in many areas such as e-learning and can also be used in mobile devices.

• 出版日期2010-3-15
• 单位中央民族大学