Critical Practices That Drive Successful Analytics and Data Projects

Six important practices can propel organizations to the forefront of their industries by tapping into the huge business value in content. Analytics tools (such as Smartlogic's software suite, Semaphore) identify, classify, extract, integrate and surface the information contained in content, making it digestible, intelligible and valuable. This kind of "content intelligence" supports existing information management, enterprise search and business intelligence systems.

Here are some words of advice based on our experience:

1. Avoid unplanned big content.
Growing volumes of data within organizations need growing resources—human and capital. Understanding what is in your unstructured data allows you to discover,  manage, control and store only the content you know is valuable. Content intelligence adds a layer of machine-readable knowledge by describing what the content is about and automatically applying metadata with that information. This metadata, in turn, gives content management systems and enterprise search and process engines context. Because the context of a piece of information determines its usefulness, relevance and treatment, you can use this metadata to keep, store and retrieve the content you know is valuable.

2. Information architecture extends the reach of your point solutions.
A strong information architecture decreases the need for information management point solutions, increases efficiency by reducing duplicate work effort and provides consistent results for end users. Content intelligence adds a semantic layer to content, ensuring that existing information management investments can all benefit from one implementation. Any system connected to enterprise content can use the facilities of a content intelligence platform to add:

• Classification facilities that increase search and findability in SharePoint and other content management platforms;
• Precise metadata that powers the lifecycles of content and records management systems;
• Enhanced indices of quality facets for precise recall by enterprise search engines;
• Rules based on meaning that drive workflow and document governance;
• Documented, open and standards-based interfaces such as RDF, OWL, SKOS and SPARQL that are key to easy integration, vendor independence, and delivery of world-class performance over enterprise volumes.

3. Precise classification delivers quality linked data.
Semantic Web projects rely heavily on manual tagging, which is inherently inconsistent even among people trained in the same field. W3.org describes this as "the chaotic, informal and weakly structured world of social approaches to information management, as exemplified by social tagging applications." Content intelligence adds structure around traditional enterprise content so that related content consistently links to like content within and outside of organizations. It automatically adds quality links to data that enrich content beyond simple descriptive tags and categorizes each piece of information as a specific type. This allows the information management tool chain to deliver on the promise of semantic data.

4. A modular platform delivers quick wins.
A modular platform that allows licensing of separate constituent parts yields flexibility and budgetary pragmatism.

• Model management and governance benefit from ontology management;
• Text mining is useful for quick-fire model building;
• Transparent, understandable and accurate metadata tagging stems from rule-based classification;
• A user experience engine ensures information is surfaced in context;
• An application framework embodying best practices means projects are off the ground in days; and
• Out of the box integrations with popular systems, such as SharePoint, ensures the legwork is taken out of systems integration.

5. Auto-classification powers text analytics.
Content intelligence is powered by a combination of a precise rules-based engine, a statistical-based natural language processing engine, and an ontology management tool. The combination delivers the quality of a rule-based system with the speed of statistical text analytics—all based on a set of rules established by each organization. The resulting standardized metadata drives the ability to find unknown information, patterns and connections within text as well as the ability to search for connections among concepts. It can be exported as linked data and used for:

• semantic processing;
• records and workflow governance;
• sentiment analysis;
• advanced content compliance;
• concept relationship mapping; and
• creating content types.

6. Content intelligence will increase ROI.
International firms, government agencies, non-profit organizations and small businesses in many industries benefit from content intelligence.

• Media companies use content intelligence to improve the quality of information feeds—boosting distribution, readership and subscriptions;
• Government authorities use content intelligence to tag information according to their standards for compliance, intelligence processing and citizen self-service;
• Healthcare companies use content intelligence to boost the level of Web self-service and improve the quality of critical health information they provide to patients;
• Investment banks use content intelligence to consolidate their information costs, better promote their primary research and automate information compliance;
• Online directories use content intelligence to increase their advertising revenues;
• Corporate intranets use content intelligence to boost use and maximize return on information assets; and
• Information managers use semantics to manage taxonomies, and ontologies, classification schemes and records retention policies.