Managing Documents: Classification and Clustering

These flashcards focus on key concepts related to document classification and clustering as discussed in the lecture.

What is the primary purpose of text classification in document management?

To categorize documents based on content for easier retrieval.

What does it mean when we say a task is 'supervised' in machine learning?

It means that the task uses labeled data to guide classification.

What is clustering in the context of document management?

Grouping documents based on content without using pre-defined labels.

What is the difference between supervised and unsupervised learning?

Supervised learning uses labeled data, while unsupervised learning does not require any labels.

What is K-means clustering?

A method that partitions documents into K groups based on the closest centroid.

What is a unique identifier in document management?

A distinct value assigned to each document to allow for precise retrieval.

What kind of classification creates multiple labels for each document?

Multi-label classification.

What is the main feature of agglomerative hierarchical clustering?

It allows for overlapping clusters and creates a hierarchy through a dendrogram.

In K-means clustering, how is the centroid of a cluster determined?

It is calculated by taking the average of all member distances from the center.

What is the output of a binary classification task?

A predicted label indicating whether a document belongs to a specific class.

What makes semi-supervised learning unique?

It combines a small amount of labeled data with a larger set of unlabeled data.

What aspect of documents do both text classification and clustering rely on?

The content of the documents.

What is the goal of the single link method in hierarchical clustering?

To find the minimum distance between two clusters.

How many clusters are generated if K is set to 5 in K-means clustering?

Five clusters.

What algorithm might take advantage of only a portion of labeled data?

Semi-supervised learning.

What is the role of machine learning algorithms in document classification?

To learn patterns in the data and predict classes for new documents.

How does one assess the similarity of documents in unsupervised clustering?

By computing the distance measures between the documents.