Implied Intelligence, LLC has developed business databases in multiple countries/markets (approximately 50). In the US as an example, our database consists of approximately 8.7 million business records, each containing home-page URL; address; keywords representing the company's products and/or services; and contact information. We also have developed chain/branch/franchise location data, and we offer enriched fields such as opening hours. We update every record and attribute in our databases monthly and we are generating approximately 240,000 new US records per month. One additional point: we have consistently found that when comparing our data to other databases, approximately 25-30% of our records represent entirely new businesses (i.e. records that are not contained in the databases of our clients). All of our data has been exclusively developed and refined by intelligent web-based mining and extraction techniques using our expertise in the domain of computational linguistics.
In addition to our existing web-based data, we also offer data mining/extraction services relative to validating, updating, and enriching existing data elements held within a database of company profiles. Another service Implied Intelligence provides is to validate and append SIC codes and/or other classifications/category heading to records. This is accomplished using a couple of techniques (1. analysis of the semantic content contained in a company name; and 2. using keywords extracted from company home pages).
Finally, apart from our Web-extracted business data which typically consists of business name, address, telephone, category and URL, we also extract company and contact information from social media sites. This data also contains fields such as date of birth, number of friends, list of friends, occupation, hobbies, previous residence and several more.
What we see as unique selling points is the combination of our linguistic and text-mining background, very large data resources and novel techniques to develop and maintain extraction algorithms. It is really the combination of having a large-scale database of high quality global websites (just business homepages, no spam), ways of rapidly extracting information out of these sites, large semantic resources and dictionaries in many languages, and finally experience and expertise in data aggregation and cleansing that creates our USPs.