About T-Rank

Introduction

T-Rank is a software provider that delivers products and solutions for network analysis. The T-Rank technologies improve both the user experience and the profitability of services like online community services, business catalogues and enterprise search (network analysis used as relevance technology).

The core of T-Rank is the set of T-Rank network analysis technologies developed by two research scientists at Telenor Research & Innovation (two of the founders of T-Rank). A network in this context is a generalized concept, and not a data network.

The T-Rank technologies can work on several types of networks:

  • A social network is a network where the members are people, and the links between the members represent social relationships (friendship) or social interaction (communication). A typical social network is the customer base of an internet community service.

  • A business network is a network where the members are people or companies, and the links between the members represent ownership in the same companies or board membership in the same companies. A business network can be built using public registers like Brønnøysund-registrene.

  • A document network is a network where the members are documents, and the links represent thematic similarity. The T-Rank technology computes the similarity between documents, and creates a network of links between the documents, which are based on the similarity scores. T-Rank can compute a document network based on most enterprise search document sets.

The networks given above are just examples. T-Rank technologies may be applied to a wide variety of data sets, whenever the data may be represented as a network or a combination of networks.

The T-Rank technologies have three main functions:

  • Ranking
  • Clustering
  • HotSpots

The technology ranks all the members of a network based on their position in the network (the network centrality). For each network member, we compute the strength of his network position. In a business network, the ranking will give each person an authority. “How important is a person in the business world, based on his business network?”

Clustering involves identifying substructures in the network, or, in other words, groups of members that are closely interconnected. In a business network, a cluster will be a group of people that own many companies together or are member of the same boards. In a social network, a cluster will be a group of people that are friends with each other and communicate a lot with each other. In a document network, a cluster represents a well-developed theme which is present in the document set.

In HotSpots, we combine network analysis with some independent quantitative parameters describing the members. In a business network, for example, we can combine network analysis with a quanititative performance measure for each company. Companies in a HotSpot will tend to rise or fall together, because of their close interconnections. Similarly, for individuals, one can define HotSpots using (for example) a measure of each person's affinity for a given product. In this case a HotSpot represents a well connected group of individuals, where the whole group is likely to adopt the product together (or not). The HotSpot approach has been tested and shown to have high predictive power.

T-Rank can deliver complete solutions based on network analysis, as software modules or as a webservice (software as a service). The T-Rank team consists of experienced professionals with over 10 years' experience in network analysis, information retrieval, and search technology. Competences represented in T-Rank span the range from research and development through challenging practical implementations and user interfaces. Our unique combination of technology and competence enables T-Rank to deliver complete solutions, which are useful for a broad spectrum of data analysis challenges--wherever a network point of view may give new insight.

The company

The T-Rank technologies have been continuously developed since 2003.

The company T-Rank AS was established in 2006, as a spin-off from Telenor R&I, and with the aim of exploiting these new technologies. Currently the company is owned by the employees.

The CEO of T-Rank is Arne Petter Omholt. Arne Petter Omholt is one of the founders of T-Rank, and has been with the company since the start. Before T-Rank, Arne Petter Omholt was CTO of the mobile services company Aspiro. He is also one of the founders of Basefarm AS, and has been CTO of Scandinavia Online. Arne Petter Omholt holds an MSc in computer science from the University of Oslo.

Svein Parnas is the Chief Technology Officer. Svein Parnas is one of the founders of T-Rank, and has been with the company since the start. Svein Parnas is an experienced professional, with search as a focus area. He has previously worked with Retriever, Basefarm and Scandinavia Online. Svein Parnas has a Master's degree in Library and Informaiton Science from Høgskolen i Oslo.

T-Rank has currently (May 2008) 6 partners: the CEO and CTO, two in development, and two in research and analysis.