HighTech Finland › Information & Communications › All articles in this section   ›  Translating data into knowledge and action

Mobility & Networking
Software & Systems
All articles in this section

 

Translating data into knowledge and action

An expert in modern linguistics and software technologies, Connexor helps its customers and partners – such as software houses, service providers, system integrators, resellers, and research laboratories in over 30 countries – find, organise, extract, and analyse the information they need from masses of documents. And put that information to work for them.

The mass of data available today can easily be more of a curse than a blessing. The less time you have, the bigger the problem of information overload is likely to become.

With the growing volume of electronic articles, blogs, and other written data out there, people are increasingly likely to overlook some of the information they need if data monitoring is only done manually. They often have more urgent things to do, and can find it difficult, in any case, to process large volumes of text and similar input.

Trying to deal with ever-larger volumes of data has encouraged people to try and organise and categorise data automatically. But is it possible to tell what texts are about without actually reading them?

Analysing attitudes

The linguists and computer scientists at Connexor have developed content analysis software to address just this problem and identify and organise relevant information from large masses of documents.

Positive and negative comments relating to two hotels generated by a Connexor analysis arranged along a timeline.

Connexor’s software goes beyond simply analysing content; it finds all the documents relevant to a specific issue, analyses what the texts are about, what they say about the topic or topics in question, and even identifies the authors’ opinions and the overall tone of the material.

This approach is based on semantic analysis and is designed to generate ambient information about specific entities and circumstances – be they companies, people, or product names, or events and relations between people and events, and whether certain entities are linked to positive or negative perceptions.

What do they really think about us?

This kind of automated analysis opens up numerous new opportunities.

Reviewing and analysing customer opinions can be very useful, for example, if a hotel chain wants to improve the services, amenities, furnishings, or other aspects of its offering.

Connexor’s automated ambience analysis can help here, as it can monitor all the positive and negative comments contained in customer feedback, news reports, blogs, and other sources of information. And enable the chain to get an early warning of potential problems, identify new trends, or take action to disseminate best practices, and avoid overlooking crucial information or making major decisions without first listening to what their customers are saying.

Customer feedback about the hotels in a chain can be automatically analysed by Connexor software to highlight positive and negative comments and link them by frequency to specific aspects of customers’ hotel experience.
> Silja Huttunen
(Published in HighTech Finland 2008)