For this scenario, one is asked to partition the vertices into clusters so that the "highly connected" vertices belong to the same cluster.

Clustering large datasets is a fundamental problem with a number of applications in machine learning. Data is often collected on different sites and clustering needs to be performed in a distributed manner with low communication. We would like the quality of the clustering in the distributed setting to match that in the centralized setting for which all the data resides on a single site. In this work, we study both graph and geometric clustering problems in two distributed models: (1) a point-to-point model, and (2) a model with a broadcast channel. We give protocols in both models which we show are nearly optimal by proving almost matching communication lower bounds. Our work highlights the surprising power of a broadcast channel for clustering problems; roughly speaking, to spectrally cluster n points or n vertices in a graph distributed across s servers, for a worst-case partitioning the communication complexity in a point-to-point model is n s, while in the broadcast model it is n + s. A similar phenomenon holds for the geometric setting as well. We implement our algorithms and demonstrate this phenomenon on real life datasets, showing that our algorithms are also very efficient in practice.

The first blackboard system was the HEARSAY-II speech understanding system (Erman et al.,1980) that evolved between 1971 and 1976. Subsequently, many systems have been built that have similar system organization and run-time behavior. The objectives of this article are (1) to define what is meant by "blackboard systems" and (2) to show the richness and diversity of blackboard system designs. The article begins with a discussion of the underlying concept behind all blackboard systems, the blackboard model of problem solving. In order to bridge the gap between a model and working systems, the blackboard framework, an extension of the basic blackboard model is introduced, including a detailed description of the model's components and their behavior.

Clustering large datasets is a fundamental problem with a number of applications in machine learning. Data is often collected on different sites and clustering needs to be performed in a distributed manner with low communication. We would like the quality of the clustering in the distributed setting to match that in the centralized setting for which all the data resides on a single site. In this work, we study both graph and geometric clustering problems in two distributed models: (1) a point-to-point model, and (2) a model with a broadcast channel. We give protocols in both models which we show are nearly optimal by proving almost matching communication lower bounds. Our work highlights the surprising power of a broadcast channel for clustering problems; roughly speaking, to spectrally cluster n points or n vertices in a graph distributed across s servers, for a worst-case partitioning the communication complexity in a point-to-point model is n · s, while in the broadcast model it is n s. A similar phenomenon holds for the geometric setting as well. We implement our algorithms and demonstrate this phenomenon on real life datasets, showing that our algorithms are also very efficient in practice.

This wolk was supported in parts by the Defense Advanced Rz-search Projects under ARPA contract no. MDA 903-77-C-0322 and National Institutes of He...Ith grant no.

The objectives of this document (a part of a retrospective monograph on the AGE Project currently in preparation) are (1) to define what is meant by blackboard systems and (2) to show the richness and diversity of blackboard system designs. In Part 1 we discussed the underlying concept behind all blackboard systems -- the blackboard model of problem solving. We also traced the history of ideas and designs of some application systems that helped shape the blackboard model. In application systems, the blackboard system components are integrated into the domain knowledge required to solve the problem at hand.

The objectives of this article are (1) to define what is meant by "blackboard systems" and (2) to show the richness and diversity of blackboard system designs. The article begins with a discussion of the underlying concept behind all blackboard systems, the blackboard model of problem solving. In Section 2 the history of ideas is traced, and the designs of some application systems that helped shape the blackboard model are detailed. Part 2 of this article which will appear in the next issue of AI Magazine, describes and contrasts some blackboard systems and discusses the characteristics of application problems suitable for the blackboard method of problem solving.

The first blackboard system was the HEARSAY-II speech understanding system (Erman et al.,1980) that evolved between 1971 and 1976. Subsequently, many systems have been built that have similar system organization and run-time behavior. The objectives of this article are (1) to define what is meant by "blackboard systems" and (2) to show the richness and diversity of blackboard system designs. The article begins with a discussion of the underlying concept behind all blackboard systems, the blackboard model of problem solving. In order to bridge the gap between a model and working systems, the blackboard framework, an extension of the basic blackboard model is introduced, including a detailed description of the model's components and their behavior. A model does not come into existence on its own, and is usually an abstraction of many examples. In Section 2 the history of ideas is traced, and the designs of some application systems that helped shape the blackboard model are detailed. Part 2 of this article which will appear in the next issue of AI Magazine, describes and contrasts some blackboard systems and discusses the characteristics of application problems suitable for the blackboard method of problem solving.