Kadoa aggregates and analyses the many product reviews that circulate on different platforms on the Internet to help consumers make the right purchase decisions. To do this, Kadoa collects unstructured data from numerous internet sources.
Excluding fake reviews
The main challenge is to ensure that the data analysed does not contradict each other by giving equal weight to positive and negative statements. Finally, fake reviews should not be included in the analysis. Advances in natural language processing (NLP) now make it possible to handle large amounts of data. Kadoa uses several text- and speech-based NLP tools. They ensure fully automated data analysis and aggregation, which is not possible by human hand.
5 sources, one trusted review
To ensure that reviews are reliable, an average of five sources are analysed for each product. A large discrepancy between one source and the total sample indicates possible fake reviews, which are then investigated further. Kadoa is not dependent on advertising partners. This means that the review platform remains independent.
However, the biggest advantage for consumers is probably the aggregation of the analysed data. On other platforms, you first have to read countless individual comments to get a picture of a product. In the medium term, the involvement of the community should further increase the relevance of product reviews. Kadoa is used as a B2B and B2C platform.
”In building Kadoa, we are able to draw on Panter's broad expertise. This includes machine learning and big data consulting. On the other hand, they helped us with business case modelling, which is essential for a start-up.
Johannes EnglerCo-Founder
