Emerjent and Espresso™

Espresso™ is indeed a "search engine", but has little in common with others in the field such as Google and Yahoo, beyond the fact that a user types in a query, and results are returned from an indexed collection of documents or web pages. Espresso™ differentiates itself from these other tools in the following ways:

Data focus: millions (not billions) of (useful) documents

Instead of indexing the entire web, Espresso™ focuses on a specific set of documents from the "deep web" that are essential to tracking emerging technologies, especially those in biotechnology, chemistry, biology, and medicine. These documents, which are automatically pulled from various sources and indexed daily, include US issued patents (1976-present) and US patent applications (2001-present), peer-reviewed research papers, NIH, NSF, and SBIR/STTR research grants, grant solicitation announcements, news stories from 1000 trusted technology sources and market and web search data. Many of these documents are not even accessible from today's web search engines.

Indexing: far beyond just keywords

Because the Espresso™ data collection is smaller and more focused relative to the popular "whole web" search engines, the information pulled from these documents goes beyond the typical indexing of words and phrases. Processing of these documents includes the extraction of people, companies, and universities, followed by mapping the connections which link these entities together into a complex "social network" or "innovation ecosystem".

This is where Espresso™ really diverges from your everyday web search engine. In addition to indexing millions of words and phrases from an extensive collection of technology documents, Espresso™ also contains a "people index" currently consisting of over 5 million people. This index not only links people with documents (such as when a person was an inventor on a patent, an author on a paper, or appeared in a news story), but also maintains a detailed, up-to-date profile of each of these people , from current affiliation to amount of grant funding awarded, and who they are most closely related to in the technology universe.

Analogous to the "people index", Espresso™ also contains a "places index", currently consisting of over 1 million corporations, academic institutions, and government labs. This "places index" links documents to places and maintains a profile of each of these places, including such data as web address, physical address, a full list of people affiliated with the organization, as well as a list of closely related places.

Search: multiple paths to better results

Espresso™ is all about technologies, people, and places, from the initial query to the results presented. In order to allow for better searches without requiring the use of complicated syntax to distinguish between, for example, "Apple" the company and "apple" the fruit, Espresso™ has three search boxes, which we find strikes the right balance between ease of use and relevance of search results. For example, when a person thinks about technology, she will usually know at least one of the following: the person associated with the technology, the place (company, university, city, state, etc.) where the technology was developed, or some keywords describing the technology. For this reason, Espresso™ allows the input of any combination of these three, which can result in some very powerful searches. For example, using current search engines for the following query is nearly impossible without much patience and time:"I met a guy at a conference named Charles Wood, from University of Nebraska, and I know he does AIDS research. Tell me more." Espresso™ makes it easy.

Everyone has experienced the typical web search: an initial vague query leads to an extremely large set of results that match the query (e.g., Results 1 - 10 of about 2,660,000). The problem with this approach is that often the reason you are searching in the first place is because you are trying to learn about a new topic. Hence, at this stage you simply aren't able to conjure up a carefully crafted query that will yield useful results. The only way around this problem with currently available search tools is to embark on a time-consuming series of searches in which you gradually refine your query based on the distilled content of the previous search results, eventually leading you to the information you desire. Espresso™'s assisted search removes all of the drudgery associated with the above approach, while still giving you control over the search where it counts. Upon initiating an assisted search, Espresso™ automatically sifts through the potentially thousands of documents that match your query and generates the most important words from those documents. At this point, you can use your newfound knowledge of the topic to jump into your search results from an informed platform rather than from ignorance. As long as you know at least one search term to begin your search, the assisted search will use that word or phrase as a seed to generate several additional terms, ultimately helping you make a more informed query.

Search: self-service or full service

Whether you prefer to have full control over your searches or would rather leave that up to our experienced team of analysts, Espresso™ may be the answer for you. For those who desire instant results, both the uniqueness and enormous utility of our search interface will be quickly apparent. For those who don't have the time, they can leave a search request with our team, who will use the full power of Espresso™ to deliver a caliber of results not seen anywhere else. And of course it goes without saying that we will customize the format of the results to fit your needs, whether it be through web access to a personalized results page, a collection of PDF documents, or a file folder with actual printed documents and analysis. We are willing to work with you.

Results organized: not just a pile of documents

Instead of spitting out a long list of documents for the user to scroll through and read, Espresso™ provides much more useful information: lists of people, places and documents conveniently separated into tabs, specific to the user's query, and ranked using a proprietary ranking algorithm. Because of the extensive data pre-processing unique to Espresso™, the software understands that documents should be ranked not only on keyword relevance but also on the "technology impact" of the actual document, and that this means different things for different documents. For example, while a patent should be ranked based on the amount it is cited by other patents, the specific section of the document in which the keyword matches appear (e.g., claims vs. description), and the actual "reputation" of its inventors, the score for a research paper needs to take into account the prestige of the publishing journal. Moreover, the ranking algorithm goes beyond documents. This means that Espresso™ grasps the concept of a person or a company and actually ranks them according to their contributions to their field. They are not just more keywords in a sea of text.

 

Sharing results: no one works in a vacuum

Because Espresso™ is a web-based application, once you have your results, so can everyone else. A simple click of the Share button allows you to give instant access to your colleagues, ideal in today's collaborative world. One group for which this feature is particularly ideal is IP attorneys, where easily sharing knowledge with colleagues or inventors should be taken for granted, but unfortunately is not. Espresso™ has the potential to change the status quo in this area.