June 7, 2010 – In a client meeting last week, we heard, “I needed to retrieve the Wilson proposal but I had to wade through two screens of results before I could locate the document I wanted.” We took a quick look and saw that the document creator had assigned uncontrolled free text index terms. The assigned terms were “cement deal” and a state request for proposal ID number. The content management system identified the name of the document creator, the date of creation, the file type, and indexed the words in the document. The user wanted to locate the “Wilson proposal” and did not know the “cement deal” term or think about searching by RFP number.
I thought about this common situation when I read a news story published on Real Wire. The write-up had a title that grabbed my attention: “OSS/BSS Data Misalignment Is Causing Havoc for CSPs.” To be honest, I had no idea what a BSS or CSP was. I knew that OSS had something to do with either information or open source. The grabber for me was the phrase “data misalignment.” I dived in.
First, OSS is the acronym for a company called Ontology Systems. Wrong about the open source, but right about the information referent. Located in London, the company has software that “generates a virtual unified model of the key relationships that define important global concepts such as customers, services, and network infrastructure.” According to its Web site:
Semantic reuse of existing data makes it possible to get results in “integration intensive” areas like Service Management, Revenue Assurance, and Data Alignment in timescales varying from 3 weeks to 3 months.
The company defines ontology as a computer readable, structured, declarative representation of concepts and their relationships. We think this is a useful definition. The idea is that the ontology allows connections across existing, dirty data in silos and systems. The method can be used in military, bioinformatics, and data federation. The company delivers a search solution!
The acronym CSP puzzled me. After some noodling, I decided that CSP was an acronym for a “customer service provider.” One clue appeared in an Ontology posting on its Web site where there were references to “service providers”. The write-up used the phrase “service provider” and then enumerated ISP, CSP, and MSP. I think defining acronyms is helpful.
Now I was ready to tackle the phrase “data misalignment.” After reading more about Ontology Systems’ approach, I decided that “data misalignment” is related to normalization of field names and other types of clean up work that is required when transforming content or processing unstructured documents into structured documents. There are a number of different methods, but regardless of tactic, some housekeeping is usually required.
The OSS approach is to focus on the problems data misalignment causes. Examples include resolving a billing dispute, identifying “stranded assets” (which I don’t fully understand), and performing audits.
I noted the discussion of data alignment, which is the flip side of data misalignment. Ontology Systems identifies these three examples. First, unifying a topology. The idea is to apply the firm’s method to a network infrastructure in order to get insight into what connects where. Second, the firm’s approach allows for “CRM Unification”. The idea is that “divergent representations of customers and products” can be reconciled. The idea is that the Ontology Systems’ approach allows a “360 degree customer view”. Finally, data alignment can help assure “service alignment.” I liked the phrase “You get what you test.” The notion is that the Ontology Systems’ method “is the first system capable of actively monitoring the alignment of services to enforce continuous quality across the service provider infrastructure.”
Okay, now I have figured out what Ontology Systems has as its focus. Let’s look at the survey report. Some of the data are germane to our use of industry standard methods to crack tough problems in building controlled term lists, ontologies (dare I use the word without the more complex connotations), and indexing of information objects.
There were four key findings that I saw germane to Access Innovations’ business.
First, in the survey sample, 96 percent of the respondents suffer operational or financial impacts from misaligned data.
Second, data misalignment occurred in accounting, which has an impact on the firm’s finances.
Third, a stunning 37 percent of those participating in the study spend more than $500,000 on data remediation.
Several observations occurred to me as I worked through the research results:
- Indexing and tagging not only helps users find information, there are broader business implications as well.
- The amount of money firms appear to be spending on data alignment was higher than I knew.
- The terminology for writing about indexing, ontologies, and data is very confusing.
We want to bring clarity to what we do, and we will try to avoid misleading coinages, undefined acronyms, and business school jargon. We know indexing and we know that our methods are world class. Our solutions improve findability but deliver the benefits documented by the survey.
President, Access Innovations