Information Transparency and Retrieval
Poor search results can increase costs as well as organizational risk, affecting efficiency and productivity. A core component in all Concept Searching solutions is the ability to automatically generate compound term metadata and elimination of end user tagging.
Identifying relevant concepts during a search effectively removes ambiguity, and enables the retrieval of information based on the specified concepts and phrases. This increases both precision and recall – two key performance measures for search accuracy and tuning.
The rich, meaningful metadata is auto-classified to one or more taxonomies, and is also fed to the search engine index. In the past, organizations have invested in content management, search, and portal technologies, but end users still struggle with finding the right content, at the right time, in the right context. Concept Searching technology addresses that issue. The end result is a taxonomy that provides a consistent infrastructure that can be shared across different applications and business divisions, through a shared enterprise metadata repository.
Improving Search Results Because Lives Depend on Them
For most organizations, search does not always deliver optimal information transparency and information retrieval. This healthcare and research firm, with offices in over 100 countries and major research centers in the UK, USA, Spain, Belgium, and China, selected Concept Searching to assist with analyzing research documents. The ability to automatically extract concepts and build an index of organization-specific terminology was recognized as a unique capability that would present the most useful documents to end users.
The search challenges faced included:
- Too many search results, lacking in relevance
- Internal content was mistagged, filed inconsistently, and unable to be found
- Inability to reuse and repurpose content previously created
- Normalization of vocabulary required
- Not possible to search on concepts, topics, subjects, or phrases
Leveraging its SharePoint investment, the company selected the conceptClassifier platform to automatically generate multi-term metadata, eliminating end user tagging. With the ability to create multiple taxonomies, it was able to auto-classify content to one or more nodes, as well as multiple taxonomies, providing the flexibility needed for the knowledge repository and information retrieval.
The easy-to-use solution, designed for subject-matter experts, enabled this organization to manage and maintain taxonomies, with minimum training.
Sharing Knowledge to Build a Better Product
This healthcare and research firm is well known for developing pharmaceuticals. As with any pharmaceutical or life sciences company, thorough research is required both initially and throughout the drug development cycle. With a global workforce of 75,000, it needed to improve information retrieval within its knowledge library, and provide the most relevant information to research professionals.
The challenges to be addressed included:
- Share expertise and knowledge across a global environment
- Help end users find, use, prioritize, and share information more quickly and easily
- Securely manage content in a policy-compliant manner
- Normalization of complex terms and vocabulary disambiguation
The organization selected Concept Searching solutions due to their automatic semantic metadata generation and superior classification capabilities. The generation of metadata eliminated end user tagging and filing. Accurate classification reduced the number of hours needed by taxonomy administrators to manage taxonomies.
Global Knowledge Library
The conceptClassifier platform provides technologies for organizations that want to proactively manage their content. The enterprise framework consists of three core capabilities – automatic semantic metadata generation, auto-classification, and taxonomy management.
This is the only solution that delivers automatic identification and extraction of concepts from within content as it is created or ingested, providing intelligent auto-classification, and enabling enterprise-class taxonomy management fully integrated with both the SharePoint Online and On-premises server environments. It runs natively in the SharePoint Term Store, with synchronous, real-time updates.
The conceptClassifier platform is optimally delivered as a complete platform with all standard features included. The intelligent automated classification and taxonomy management provide organizations with the ability to develop a consistent structure, to more effectively manage their content assets and utilize these assets to improve business processes.
The solution expedites the taxonomy development process, by providing the rich, automatically generated multi-term metadata to administrators through innovative auto-clue suggestion, based on the concepts within the content.
Translating the outcomes from clinical trials plays a vital role in the care, and the life expectancy, of patients, when traditional, standard care treatments are not as effective.
The successful outcome of a clinical trial must answer the question of how to help consumers, clinicians, purchasers, and policy makers to make informed decisions that will improve healthcare, at both individual and population levels. In the US, rigorous approval by the Food and Drug Administration is a mandatory component that can offer patients new hope for better health, regardless of whether for a trial, vaccine, or medical device.
Using Concept Searching’s high-performance semantic search engine, conceptSearch, a unified view of a clinical trial can be achieved. The ability to integrate diverse repositories and applications provides capabilities for clinicians to achieve granular or broad views of a trial, as well as track the associated compliance and information governance challenges from a single, collaborative project view. Concept Searching’s unique compound term processing capabilities ensure legal protection for participants, including HIPAA and data privacy, as content is created or ingested.
Over 7,700 Clinical Trials and Still Going Strong
Having successfully conducted over 7,700 clinical research trials, this North Carolina company is one of the top tier clinical research sites in the US. From an information technology perspective, clinical trials consist of highly structured and unstructured data. Together, these data types must ultimately offer a comprehensive, integrated set of conclusions and outcomes, which takes into account both the complexity of structured data and the human insight derived from unstructured content.
What were the information technology considerations for the solution?
- Access to unstructured, structured, and semi-structured data, providing a holistic view of all data points, regardless of format
- Semantic consistency, common nomenclature, and vocabulary disambiguation
- Multi-tier, hybrid search approach
- Secure collaboration
- Legal protection for participants – HIPAA, data privacy
- FDA compliance and information governance
Concept Searching is able to offer a unique solution with the ability to integrate structured, semi-structured, and unstructured content, to give a single view. As an enterprise technology solution, it enables organizations to address the various challenges associated with clinical trials.
Content Optimization – Removing Obsolete and Redundant Content, to Make Information Retrieval Relevant and Accurate
This pharmaceutical company has a portfolio of analgesics, to address pain and a variety of devastating and hard-to-treat brain, autoimmune, and inflammatory illnesses. In its preparations for migration to SharePoint, Concept Searching solutions helped the content optimization process.
Its challenges included:
- Redundant documents, different versions, revisions, and near-duplicates were impacting search outcomes, causing irrelevant and inaccurate results
- Ensuring all records in the corpus of content had been declared
- Ensuring all data privacy and confidential information was protected with the appropriate security before the migration
- Reducing the number of servers required
Using the conceptClassifier platform and conceptTaxonomyWorkflow, the organization was able to reduce the quantity of documents that no longer had value. By doing so, it was able to expedite the process of migration.
In addition, after the migration, it benefited from an improved search experience, as irrelevant and out-of-date documents had been removed, so were no longer available during the search and retrieval process.
Some refer to it as deletion. But it’s more than that. What about that one document that contains value? Don’t you want to keep it?
Content optimization is the process of removing information from active systems, through deletion or archiving, and identifying duplicates, near duplicates, undeclared records, and data privacy exposures. It eliminates obsolete and trivial content, which has been saved but is of no value.
Concept Searching recommends that content optimization is done on a quarterly basis. It is estimated that 69 percent of an organization’s content can, and should, be deleted.
Did you know?
- 25 percent of content is duplicate
- 10 percent has no business value
- 90 percent of documents are never accessed after creation
- 65 percent of documents are accessed only once
And the risks are:
- PII, PHI, and PCI data breaches
- Uncontrolled intellectual property
- Unmanaged documents of record
- Unsecured confidential company information
Content optimization not only reduces storage but dramatically improves information retrieval, as it eliminates false positives and irrelevant information from retrieval results.
Migration projects typically have questionable success rates. You are migrating a vital portion of your business. Are you willing to accept failure?
Occurring after content optimization, intelligent migration identifies the content to be moved, and an administrator defines where it will be moved. Concept Searching technology generates concepts before the actual migration. Temporary taxonomies are used to refine the content by concept, or client, or topic, within specific taxonomies. This effectively groups like content, so when migrated it retains the likeness or similarity of related documents.
An intelligent approach to migration can be achieved. As content is migrated, it is analyzed for organizationally-defined descriptors and vocabularies, which will automatically classify the content to taxonomies, or optionally to the SharePoint Term Store, and automatically apply organizationally-defined workflows, to process the content to the appropriate repository for review and disposition.
This has an added benefit after migration, improving information retrieval and providing the taxonomy hierarchy to end users as a navigational aid, offering like topics in the taxonomy for selection and exploration, refining their initial queries.
Performing Content Optimization Guaranteed a Smooth Migration, Content Was Organized, and Search Outcomes Were Significantly Improved.
This US-based pharmaceutical company has a suite of analgesic products. In moving to SharePoint, it was looking for a way to facilitate the migration and ensure accurate processes.
The challenges included:
- Enrich remaining content with multi-term metadata
- Ensure migrated corpus of content was organized and easily managed
- Thorough planning for migration, from SharePoint On-premises to SharePoint Online
- Clean up the corpus of content, and identify undeclared records, data privacy and confidential information exposures, and noncompliant instances through content optimization
Selecting Concept Searching solutions, the company performed first content optimization, then migration. The amount of content to be migrated was greatly reduced after the content optimization. But it was still beyond the scope of human intervention to read and correctly categorize all remaining documents.
Using the automatic multi-term semantic metadata generation and auto-classification capabilities, it was able to accurately classify the remaining content to appropriate taxonomies. This ensured content was organized by subject after the migration, and so more rapidly found during search.