PubMed Query Log Analysis and Use in Access Inhancement

生物医学文献检索是获取越来越多信息的主要切入点。PubMed/MEDLINE是用于此目的的最广泛的服务。然而，在PubMed中找到与用户信息需求相关的引文并不总是那么容易。提高我们对不断增长的PubMed用户群体、他们的信息需求以及他们满足这些需求的方式的理解，为改善PubMed提供的信息服务和信息获取提供了机会。用于理解和描述搜索引擎用户特征的一个资源是事务日志。我们通过一个月的PubMed日志调查用户交互，重点分析用户需求、查询的不同方面（如长度）和用户搜索习惯。在查询日志分析的基础上，我们进一步开发了查询公式中的自动搜索辅助工具，即相关查询（RQ）。RQ侧重于发现包含初始用户搜索词的流行查询，目的是帮助用户以更精确的方式描述他们的信息需求（即相对于用户输入增加专一性）。自2009年1月以来，这一援助已被整合到PubMed中。使用点击量数据进行的自动评估显示，每天新功能的使用率在6%到10%之间，这表明它已经迅速成为PubMed中一个受欢迎的新功能。受其成功的启发，我们目前正在试验其他最先进的方法，以进一步提高查询建议的质量，并扩展到查询规范之外。此外，我们正专注于开发计算技术，以响应PubMed中返回零结果的查询。正如我们的日志分析所示，大约15%的PubMed搜索属于这一类。在某些情况下，确实没有文档或摘要可以满足特定的查询。然而，在分析提交到PubMed的一个月的查询时，我们发现，通常情况下，没有检索到结果的查询如果构造不同，就会检索到相关的内容。我们目前正在识别不成功查询的一些特征，并教计算机自动学习用户在构建新查询时最常应用的更改。

Biomedical literature search is the main entry point for an ever-increasing range of information. PubMed/MEDLINE is the most widely used service for this purpose. However, finding citations relevant to a users information need is not always easy in PubMed. Improving our understanding of the growing population of PubMed users, their information needs and the way in which they meet these needs opens opportunities to improve information services and information access provided by PubMed. One resource for understanding and characterizing patrons of search engines is the transaction logs. Our investigation of user interactions through one month of PubMed logs focused on analyzing user needs, different aspects of queries (e.g. length), and user search habits. Built on the query log analysis, we further developed an automatic search aid in query formulation, namely Related Queries (RQ). RQ focuses on finding popular queries that contain the initial user search term with a goal of helping users describe their information needs in a more precise manner (i.e. increase of specificity relative to the user input). This aid has been integrated into PubMed since January 2009. Automatic assessment using clickthrough data shows that each day, the new feature is used consistently between 6% and 10% of the time when it is shown, suggesting that it has quickly become a popular new feature in PubMed. Inspired by its success, we are currently experimenting with other state-of-the-art methods for further improving the quality of query suggestions and expanding beyond the query specification. In addition, we are focusing on developing computational techniques in response to queries that return zero results in PubMed. As shown in our log analysis, about 15% of PubMed searches fall into this category. In some cases there really is no document or abstract that will satisfy a particular query. However, in analyzing one month of queries submitted to PubMed, we find that more often than not, queries that retrieved no results are queries that would retrieve something relevant if they were constructed differently. We are currently identifying some of the characteristics of unsuccessful queries and teaching computers to automatically learn the changes that users most often apply in constructing new, corrected queries. Not only can log analysis help PubMed search as a whole, it can play an important role in developing tools for improving the links between different Entrez databases. Through our analysis of PubMed logs, we learn that people search certain biomedical concepts more often than others and that there exist strong associations between different concepts. For example, a disease name often co-occurs with gene/protein and drug names. As a result, we have worked towards the development of different PubMed sensors: automatic tools for recognizing biomedical concepts and building links to related data outside of biomedical literature. This will allow PubMed users to be more readily drawn to related data that could lead to serendipitous discoveries.