Ideas & Players

Move into uncharted territories

Searching and tracking known ideas and players is one thing. But Mergeflow helps you discover ideas and players that are relevant but that you didn't even know might exist.

Discover and track...

People

See who the scientists, entrepreneurs and inventors are behind the technologies that matter to you. Identify potential collaboration partners for your next project, and experts who can bring valuable new ideas to your organization.

Organizations

Discover companies, R&D organizations, and investors within and adjacent to your fields of interest. Identify those organizations that drive the next generation of technologies and business models.

Concepts

Mergeflow automatically assigns patent classes to all documents that are not patents. Patent classes are a wide-coverage and internationally standardized concepts system. Having them available for all documents, not just patents, enables you to go beyond simple keyword searches.

Materials

Using cutting-edge machine learning methods, Mergeflow discovers chemical named entities in documents. This helps you discover previously undocumented compounds, and materials with certain properties.

In addition to people, organizations, concepts,and materials, Mergeflow also detects names of events, investors, locations, patent classes, and various medical entities, such as organisms and diseases.

See all entity types detected by Mergeflow →

In combination with discovering markets and investments and emerging technologies, Mergeflow's entity detection helps you go from sifting through documents to exploring new topics and ideas.

Start discovering

7-day free trial. Full access to all Innovator Plan features. No credit card, no commitment.
Please fill out all fields.*
* We will not sell your information to marketers and will only collect and use your information in accordance with our Privacy Policy.

Powerful analytics

Mergeflow uses various natural language processing, machine learning, semantic modeling, and other methods for discovering companies, people, materials, topics, and other categories in text data.

Differentiated approach

There isn't the one analytics technology. For example, an algorithm for discovering people names works quite differently from one that spots company names, or an algorithm that assigns topics or patent classes to whole texts.

See how players relate to each other

In order to judge the relevance of a player, it is helpful to see how this organization or person relates to others. Mergeflow uses various types of interactive visualizations to help you uncover these patterns.

Discover central players

Co-author or inventor networks from science publications or patents help identify central people in a technology field or at a company. Mergeflow plots such networks, and draws lines between any two objects that appear in common contexts.

network of inventors at Google and DeepMind working on reinforcement learning

Explore relations between topics and players

"For my search on ‘smart city’, I want to see which emerging technologies are connected to which companies."

Mergeflow uses Sankey charts to visualize such relations between any two types of object categories.

Sankey chart showing how various computing, network, and other technologies relate to companies such as Baidu, Bosch, Huawei, IBM, and Intel

You might also be interested in:

Markets & investments

Business data from texts, made accessible.

Discover markets & investments

Emerging technologies

24/7 disruptive innovations discovery.

Discover emerging technologies

Stay in the know

Continuous discovery, delivered to your inbox.

Discover Weekly360s
© Copyright 2020 Mergeflow AG.
Company Details
Privacy

Chemical named entities

Chemical named entities are names of molecules that are mentioned in texts, for example hydrogen peroxide or LiMn2O4. Mergeflow can recognize such names, including ones that it has never seen before. In order to be able to do this, we have developed a new machine learning algorithm.

Chemical named entity recognition has seen substantial advances recently. For example, here is a recent approach:

Tshitoyan, V, Dagdelen, J, Weston, L, Dunn, A, Rong, Z, Kononova, O, Persson, KA, Ceder, G & Jain, A (2019). Unsupervised word embeddings capture latent knowledge from materials science literature. Nature 571, 95-98. → Get the paper


Undocumented compounds

By 'compound' we mean molecules that are made of atoms from different elements. For example, water (H2O) is a compound because it is made of different elements, hydrogen (H) and oxygen (O). By contrast, hydrogen gas (H2) is not a compound because it is made of only one type of element (H).

When publications describe new, previously undocumented, compounds, Mergeflow helps you discover these compounds. This is because our chemical named entity detection algorithm is built such that it can also recognize new compound names that it has never seen before.