Data visualization: How to play large data

Although visualization is not the most challenging part of the data analysis field, it can be said to be the most important aspect. Of course, storage, database query processing, and algorithms are all very important--visualization is not possible without them--but in a data-driven world, they are just at the base level. There are 6 startups that are trying to fundamentally change the visualization of the data. Some of these are highly complex visual processes, some not. Although none is perfect, what they do will make us rethink: what does the data mean?

Ayasdi

AYASDI from the Indian language, is "looking for" meaning. Gurjeet Singh at Stanford University, Gunnar Carlsson and Harlan Sexton have been working to apply the research methodology of topology to data analysis. In 2008, they jointly established the AYASDI company. When AYASDI was founded, he received 3.5 million dollars of support from DARPA, the US Defense Advanced Research Project group.

The bottom layer of AYASDI uses the HBase data storage, then uses the topological data analysis technology and hundreds of machine learning algorithms to process the complex dataset, and finally determines the similarity between the data nodes. For the end user, this looks more like a dataset topology, but emphasizes the associated data points in the cluster.

AYASDI's technology has an important feature, unlike other systems that require similar search-query statements, AYASDI can automatically discover hidden patterns from the data. One of AYASDI's applications is in the field of medical research, and Eric Schadt, director of the Genetics and Multi-scale Biology department at Sinai Medical School, led a team that used AYASDI techniques to study genetic predisposition for some diseases, and using AYASDI data analysis techniques, Helped to find 14 variants of breast cancer.

BeyondCore

In fact, the basic premise of beyondcore operation is the same as that of AYASDI-just showing the user a significant connection without considering how the result was discovered-although BeyondCore uses standardized charts, But in fact there is another option: when the software discovers the relevance result, it lets the avatar (see picture) inform the user.

ClearStory

ClearStory has a unique product, even though many details of the product are not published. Perhaps only after the product is officially released can we "see uncover". ClearStory will be connected to data stored in databases, networks, and other sources, dynamically updating when data sources change, and adding virtual data mining components to enable users to analyze data and understand information. In March this year, ClearStory won the favor from Google VCs, Andrison Hollowitz, Khosla Ventures and other VCs, with an unknown amount of funding.

DataHero

DataHero was founded by Chris Neumann, the backbone of the big data analysis business from Aster, and BMW's senior engineer Jeff Zabel, which has received 1 million of billions of dollars in funding at its initial stage. They're trying to bridge the gap between big data and ordinary users by using the slogan "Your data is your decision," to help users analyze the data closely, and to get a better understanding of it in a visual format.

According to foreign media reports, DataHero is a browser-based platform, data output, collation and interpretation will be completed in a separate window, rather than the usual meaning of the use of different applications. Through the browser, users can input a variety of network services information into the DataHero, can also download reports from the public database, but also directly to the local electronic forms and files uploaded to DataHero. DataHero automatically categorizes and labels the information according to its algorithm.

Platfora

The start-up company Platfora claims to "Hadpoop civilians". The Company believes that only the jumbled data for effective processing, visualization, so that the data into ordinary users can understand things, "big data" can really become a business, the big data start-up company has won the 20 million dollar round B financing.

Platfora on the basis of HADPOOP data operation, and provide users with an Easy-to-use platform for operation. Because Hadpoop has a lot of different distributions, so platfora one of the key is to ensure that it can run on all distributions, which greatly reduce the hadpoop threshold, so that more people can experience hadpoop technical advantages, to achieve the real sense of "civilian".

Zoomdata

Start-up Zoomdata is one of the few data-analysis companies that support mobile devices, and their data visualization systems enable large data streams to be translated into touch-friendly, artistically dimensional data. Apple ipad and Android tablet users can zoom the data visualization interface with their fingers, and the data will be updated in real time with different levels of interface scaling. Zoomdata's data visualization technology supports a variety of data sources, including social media, enterprise applications, and Hadoop HDFS data.