Driving new digital revenues through text mining
Posted by Peter Marsh on Fri, Jan 29, 2010
Searching and browsing a website is a lot like looking for data through a keyhole. Text mining opens the door to this data and allows it to be viewed and discovered in a much wider frame of reference. As a result, this content enrichments process helps to transform simple search into true knowledge management. This leads to greater time-on-site for web site visitors, a more relevant search experience, and new opportunities for media companies to build revenue streams that monetize digital news and advertising content.
Sentiment and Semantics
Most everyone uses Google for searching. But, the problem with Google is that users are "promiscuous." They often come in and out of Google searches almost at random, without any particular loyalty to the search results or the sites they visit. They cannot easily orient themselves to the results lists they see, and they cannot determine the context in which the results are presented. In addition, Google has no idea of a user's moods or goals when conducting the search in the first place.
Most users do not know how (nor want to take the time) to construct Boolean search strings, so they rely on simple keyword or phrase searches. This type of searching is inherently inadequate, because it ignores the strengths of the editorial content. Good journalists make use of connotations, inferences, metaphors, and synonyms when crafting a story, yet these writing techniques often get lost with keyword searches. Worse yet, simple keyword searches many times yield thousands of results, which overwhelms the searcher.
Text mining helps to solve these search challenges by creating connections within the content based on associations, related themes, and cross-database references that add meaning to the content being searched. Meaning is determined by tagging, metadata, taxonomies, grammar and statistics. Text mining harnesses the unstructured, free-form nature of human language by applying these linguistics rules to help make the content more meaningful and contextually relevant. In fact, recent developments in content enrichment technology is making it possible to automatically detect the sentiment of the content - e.g. irony, sarcasm, bias - which helps to make the results even more relevant for the searcher.
Why is Text Mining Important to News Publishers?
Media companies around the globe are faced with an explosion of structured and unstructured information in today's digital world. Yet, a media company's audience members - internal and external users alike - are starving for knowledge.
By implementing software solutions designed to optimize the content enrichment process, media companies are able to provide their users with the most robust, adaptable, secure, and high-performing content experience possible.
Because of the nuances and subtleties involved in human language, a good content enrichment solution must support automatic and manual text mining, or a combination of both. In addition, the text mining and content enrichment process must be fully integrated into a media company's digital content management applications.
Top Ten List
Text mining enhances the website experience for media companies and their customers, which leads to greater marketability and new revenue opportunities. The following list represents the top ten business benefits to be derived from the text mining process:
- Enhanced search experience for external users, which leads to greater time on site and new revenue opportunities
- Increased revenues as a result of greater pages views per session
- Enhanced search experience for internal users, which leads to deeper discovery and better knowledge management
- Improved newsroom team satisfaction and productivity
- Improved product quality, which leads to increased end-user loyalty
- Greater content consistency (classifications and taxonomies) across newsrooms, along with automated workflow processes related to tagging - i.e. the ability to relate appropriate photos, videos and related articles to news stories without manual intervention
- High probability that content will be found by search engines, which leads to increased traffic and website access
- Better search engine optimization - the most relevant metadata on a page can be used to create associated naming conventions, URL's and more effective HTML tags
- Dynamic content delivery and personalization - allowing users to get the content they want, when they want it
- Increased advertising revenues resulting from contextual ad delivery and geo-referencing
Summary
Experts at IDC estimate that up to 80% of the data managed by organizations is stored in a free text format. This data holds massive untapped value, but the unstructured nature of human text makes it difficult to give users the search results they really want. A text mining component, when integrated as part of a total content management solution, helps to solve this problem by extracting metadata (concepts, entities, persons and places) and finding similar or related content. As a result, users will be assured of getting accurate search results, but more significantly, the resulting content will be put into the correct context.
There's no question that we're in the throes of an information explosion, with the media industry sitting at the center of all the action. Search technologies are great at helping us quickly find the content we are looking for. But, search engines do not help us to discover things we're unaware of. In short, search finds documents not knowledge. To harness the untapped potential in all this information, a content enrichment solution provides the technology to bring relevance, meaning, and added value to the content we consume.
For more information on how the Atex Text Mining solutions can help your media company generate new revenues, increase website traffic, and expand the important stickiness factor, please click here.