A Non-patent literature (NPL) search and analysis is an integral part of technology and innovation research. Non-patent literature analysis combined with patent research can provide valuable insights and a more comprehensive view of the technology landscape.

Unlike patent research, which benefits from the availability of many structured/indexed databases, non-patent literature is challenged by the exponential growth of information sources available worldwide.

Conducting Non-patent literature (NPL) search presents its own share of challenges:

• It is a time-consuming process, as one must first prepare a list of sources and then conduct a search in each of those sources.
• A search may be incomplete, as not all of the important sources would have been consulted.
• A lot depends on the analyst when it comes to quality.
• Setting up of alerts is time-consuming, since alerts are required for each of the multiple sources.

SciTech Patent Art has developed proprietary techniques to simplify Non-patent literature (NPL) search, keeping in mind the challenges associated with NPL searching and the dynamic nature of the web.

Our team of software engineers have developed tools and solutions, using their considerable experience in the Intellectual Property (IP) industry, thereby delivering value to technology and innovation research.

SciTech Patent Art has developed a Deep Web search tool as an example of such a solution.

As the name suggests, Deep Web searches refer to searching for information, which is not indexed by public search engines such as Google, Bing, or Yahoo.
Information that is accessible through such public search engines, which is called “Surface web”, is < 5%, whereas the Deep Web, which comprises about 95% of the web, contains information that is either buried deep within search results or cannot be retrieved for the following reasons:

• Public search engines prioritize content based on geographic location, promotions, etc., which may affect the reliability of the information.
• There may be some pages that are not indexed due to the owner’s discretion.
• There are certain websites that are private and require authentication to access.
• Many other websites require Captchas in order to prevent automated data scraping.

SciTech Patent Art’s approach to implementing the Deep Web search platform.

The Deep Web search platform built by SciTech Patent Art is a collaborative effort of software engineers, knowledge scientists and search experts at SciTech Patent Art. The team brings technical expertise and domain-specific knowledge together to develop an efficient and comprehensive platform for accessing Non-patent literature (NPL) search.

SciTech Patent Art’s Deep Web search platform is built on our strengths:

Domain expertise

– Our team of knowledge scientists curates highly targeted data sources as they have deep knowledge of technologies spanning across chemistry, polymer science, food technology, packaging, mechanical, automotive, medical devices, pharmaceuticals, biotechnology, material science, electrical and electronics, semiconductors, etc.

Search expertise –

Our team of knowledge scientists have many years of experience searching through multiple databases and sources of non-patent literature. They are well-versed in crafting creative search strategies to extract highly relevant art useful for critical search and analytics projects.

Data engineering expertise –

Successful execution of “Deep Web” searches requires core data engineering capability that is offered by our team of software engineers who possess industry knowledge and experience.

Machine learning integration –

Further machine learning-based algorithms are integrated into these domain-specific databases, which adds sophistication to the platform. There are two levels at which machine learning algorithms are developed and integrated:

• We use machine learning algorithms to develop a comprehensive topic / technology-specific synonym list based on an initial list curated by our team of knowledge scientists.
• Machine learning algorithms are also used to automate categorization of documents based on the synonym list and contextual information.

What benefits does SciTech’s approach to Non-patent literature (NPL) search provide?

Customized data solutions–

The expertise of our team of knowledge scientists and software engineers enables us to develop customized crawlers and scrapers, develop algorithms to structure data, create data pipelines, etc.

Iterative data solutions –

An iterative approach is adopted to ensure that the deep web search platform is routinely refined and updated with the latest data sources, documents, synonyms and technology information.