Addressing Unstructured Content Within the enterprise

He takes our questions


What makes you enter this market?


Semagix is entering this market to specifically address unstructured content within the enterprise and the need to leverage information as an asset by extracting meaning from this content. We also address the need to leverage existing investments in content related technologies that have failed to deliver on their promise of enhancing and improving business applications and processes.


How?

Through the creation and use of domain-specific, semantically enhanced metadata. Our product has been developed from the ground up on XML and metadata to provide enhanced information management services to third party systems via SOAP. These systems can call on and use this semantic metadata to improve search, facilitate effective collaboration and the real time sharing of information, personalise interactions to a highly granular and effective level. Semagix takes the view that the effective creation and management of metadata lies at the heart of any content or information related strategy.


Can you differentiate your proposition from existing ones?


Yes. Semagix has two main points of differentiation. First is the ease of integration into every aspect of an enterprise's content value chain, from creation to publishing, irrespective of whether the content is structured, semi-structured or unstructured. The second point of differentiation lies with the depth and quality of the metadata that Semagix’s Freedom product creates and uses.



How does the integration take place, and what advantages would it offer in comparison to other cm systems?


Firstly we do not position ourselves as a content management systems vendor per se, but rather as a technology that underpins content management systems.

One important aspect of our technology is its ability to easily integrate with third party applications, which occurs at the metadata level through FREEDOM’s metadata adapters. These manage the exchange and translation of metadata from FREEDOM’s metabase to the third party application and vice versa.

The advantage that this offers content centric applications is that from a technical perspective the content has context and meaning. This increases the value of the system through being able to retrieve, use, collaborate on and publish content that is in-context and relevant to the user’s criteria.



Please briefly describe the main processes supported by the tool


We have toolkits to design the FREEDOM ontology and the metabase.

The Ontology defines the concept classes we wish to maintain to facilitate the retrieval and linking of content. In traditional taxonomy lifecycle management terms, this could be considered to represent the taxonomy, however it is not hierarchical based and its key ingredients are a vocabulary of basic terms that holds precise relationships that define the meaning of those terms.

The Metabase stores metadata (both semantic and syntactic) related to content items in either custom formats or one-or-more defined multiple metadata formats such as RDF, PRISM, Dublin Core, and SCORM. The toolkits provide a flexible definition of the structure of the metabase so for example, we can define the 800 field PRISM metadata standard. This means that we're providing an infrastructure via metadata for an organisation to manage content through the enterprise and indeed to provide and receive information with other third parties/companies adhering to a similar structure.

The enhanced semantic metadata is created via a ‘classifier committee’ approach (initially to create the classification metadata and to provide the semantic enhancement process with context) and then via the semantic enhancement process that extracts relevant information from content, maps it onto the ontology to create additional semantic metadata to populate fields within the specific metadata record in the metabase.

We claim to differentiate thanks to our technical ability to consistently produce actionable (timely, in-context and relevant) information.



How is this ability supported by the system?


First, the system is near real time. As content is received into the organisation it can be classified and enhanced immediately and made available for browsing within seconds. Secondly, content is associated via semantic metadata with entities in the Ontology (a profile) and these can be matched to the profile of users so that the information is presented in context and in a timely fashion.

The ontology and metabase are populated and updated using agents connected to trusted knowledge and content sources.

Content sources are connected via FREEDOM's patented content agents that trawl through networks, web sites, Intranets, some content repositories and databases).
These agents work autonomously (scheduled with manual overrides) to ensure that
content is classified, enhanced and indexed, so that information can be retrieved when it's required.

Structured data is collected by Knowledge Agents. These are specifically used to maintain the domain Ontology with facts/entities (people's names, company names) and relationships between these entities. This Ontology represents the domain specific knowledge that generates part of the semantic metadata.

Bespoke semantic applications are created using the services of the Semantic Query Server. This provides mappings onto the metabase and Ontology to create any type of interface included guided browsing, real time alerting and portal type applications.

Another one of our differentiators is the ability for end users to undertake knowledge discovery simultaneously through the Ontology and metabase to better understand
the implications of why a content item might be important and worthy of
further analysis. This is also supportive of deep intelligence and risk assessment applications for example.

For example?


Anti-Money Laundering application. Typically the Know Your Customer checks carried out within legal and financial services are costly, in terms of HR and content subscriptions, and are not flexible enough to adjust to the constant changes in the area of compliance.

Compliance teams are required to have the processes and controls in place to cross reference large volumes of information from a wide variety of sources, whether these are in structured or unstructured formats. From this breadth of content compliance teams have to identify and verify a person and the risk they pose before they are processed as a customer – currently this is a timely and costly process.

Semagix’s technology is ideal for this; as it can aggregate large volumes of information and enhances this information through entity extraction and link analysis. This automates the name and address identification and verification process, whilst checking that the person does not appear on any internal or external watchlists. In addition to this huge process saving there are reductions in the content costs, as the system only uses a specific set of content fields that are relevant to anti-money laundering processes.

How is Semagix going to tackle the market?


Direct and Indirect through a selected group of system integrators

What do you perceive as the greatest challenge in finding customers?


Competition for budget across many applications

What are the geographical priorities for Semagix (UK, Europe, US or other?)


UK and US

How do you position the product in the content technologies space


FREEDOM underpins any Enterprise Content Management application (content management, document management, digital asset management) by creating content structure through rich semantic m