Four use cases for triplestores
• Search and Discovery Search and discovery covers a broad set of applications but generally refers to any use case where organizations want to improve the content delivered through search. The ability to do this with semantic search improves the results. eg... ...imagine a user searching on the words “maintenance” and “operations.” The user gains the advantage of having the system automatically also search on “repairs.” Related concepts are recommended to the user. The user is also given the option of comparing this paragraph to other paragraphs in other documents, paragraphs that are also relevant.
• Personalized Content and Recommendations A second use case has to do with targeted semantic recommendations made for both internal users and website visitors. The full text search criteria, search categories, date of the search and the log can all be maintained in the triplestore. Search actions are classified and enriched using a taxonomy of terms. In the darker blue, this semantic knowledge is used to deliver related articles which enhance the website experience. This leads to more time on the site, targeted recommendations, increased customer loyalty, revenue and advertising. Since website visitors often comment, vote, post and perform other actions, all of this can also be used in the semantic targeting applications.
• Dynamic Information Products ....what a publishing analyst would see to better understand the articles trending on their site based on key words and phrases that are part of the search. In the upper right, the analyst sees how the phrases “American Football” and “World Cup Soccer” are trending during the 2014 World Cup. They have visibility into related articles that are displayed to the readers. An analysis of the article with highlighted words and phrases is available for review. The analyst can also look at faceted search results, and quickly modify the search using related words and concepts. This provides intelligence about site performance in a very timely way. As you would imagine, since all other articles are semantically analyzed and indexed, the relevancy for the website visitor reaches heights never achieved before. All of the data is stored inside a triplestore.
• Data Visualization These RDF statements can also be used to draw a visual graph with interconnected nodes linked by the predicate of the triple. This type of analysis is especially useful in pattern detection, anomaly identification and network analysis. It shows entities linked together that simply cannot be accurately expressed in the form of charts.