Decentralized Partnership Formation

This project applies (1) automated Web-based retrieval of semantic associations between brand names and 400+ brand personality and competences indicators, and (2) decentralized k-clique matching, to develop a peer-to-peer matching system for organizations to build cooperative partnerships. Scientific aims are to test whether Web-mining consumers’ brand perceptions gives externally valid results (CommScience), and whether the matching protocol works on messy real-world data (CompScience). The practical aim is to aid small and medium sized businesses identifying meaningful partnerships.

In Year 1, the project delivered (a) a five-dimensional brand personality indicator list consisting of 400+ high-usage adjectives (English and Dutch); (b) several large data sets (400,000+ measurements) on semantic associations in different markets (Fortune 500, fashion, retailing, automobiles, entertainment); (c) a literature review considering different similarity measures operationalizing semantic associations; (d) a fine-tuned k-clique matching protocol resilient to messy, often inconsistent, Web-retrieved association data; (e) an protocol extension allowing clusters of varying size (N=2 to 5); (f) many simulations suggesting partnerships within markets.

This year, we would like to further validate and disseminate its results. To this end, we set four specific goals: (i) collect datasets on less “visible” – small and medium sized – brands; (ii) conduct two validation studies where consumers assess system-generated partnerships in terms of fit, complementarity, expertise, trust, and liking; (iii) co-write a manuscript describing our approach and results (from brand association retrieval, to k-clique matching, to consumer validation) for a ISI-ranked marketing journal; (iv) implement a demo of the actual peer-to-peer system, where organizations’ software agents exchange information to estimate mutual partnership potential.

While the first project year has been very productive in establishing the ground work, a second year should allow us to establish scientific and practical relevance, and, in addition, further integration of communication and computer sciences to advance research in both areas.