Hello, We are CYRAM Inc., a technology company specializing in Graph Data Science.

At CYRAM, we have been conducting various organizational network analyses using NetMiner with data provided by our clients. 

In this session, we will introduce the procedures and methods of organizational network analysis.

Definition of ONA

Organizational Network Analysis (ONA) is a method for visualizing and analyzing relationships and communication patterns among individuals within an organization. 

It reveals hidden networks, enabling insights into knowledge flows, fostering innovation, developing leadership, and predicting employee turnover, making it highly applicable in fields 

such as HRM and People Analytics.

Data Collection Method

1. Data Collection

First, data is collected through surveys or extracted from emails, messaging platforms, or collaboration tools.

2. Data Structure

The collected data should ultimately be organized into the following Edge List format.

Node data typically refers to employees' HR profile data. It is recommended to use unique attributes, such as employee ID, rather than names.

Organizational Network Analysis Using NetMiner

Now, let me introduce some useful analytical methods utilizing this data.

1. Analysis of Collaboration and Communication Characteristics Within the Organization

"Can we objectively compare which is more active in our organization: work collaboration or informal communication?"

Firstly, by applying algorithms from social network analysis, such as Density, Average Degree, Centralization, and Reciprocity, we can assess the structural characteristics of individual networks. 

This allows us to compare the volume, structure, and patterns of communication between the company and specific organizations. 

Additionally, within the same organization, we can evaluate whether work-related information sharing relationships are more active than informal personal relationships.

※ Analysis Methods

Here's how to analyze the structual characteristics of individual network in NetMiner:

1. Go to File > Open and locate the Org_Net_Tiny1.nmf file (Example Analysis Data: Download the attached file below.), then open it.

2. In the top menu, click Analyze > Properties > Network > Multiple.

3. Select the network from the data, then in Properties, choose # of Links, Density, Average Degree, Reciprocity.

4. Click Run Process at the bottom.

※ Analysis Results

1. work interact vs work help

- Both relationships occur during work, with work help showing slightly higher reciprocity.

- This indicates that voluntary relationships of giving and receiving help are formed when the parties involved are aware of each other.

2. personal friend and personal help

- Personal friend and personal help have identical metrics, indicating that they represent the same network.

3. personal know Relationship

- As relationships among employees within the company, the Density and Average Degree are the lowest.

- This indicates that the relationship is in an inactive state.

2. Analysis of Interdepartmental Collaboration Relationships and Roles

"Why does this department always work overtime? Can we understand and adjust the roles of each department in organizational communication?"

Secondly, by applying the Blockmodel algorithm, the network can be summarized into relationships between departments within the organization. 

This helps determine whether interactions within departments are more active than those between departments. 

These analysis results can be used to understand and adjust the positions of departments within the overall network. 

Additionally, this method can be adapted to summarize and analyze the entire network in terms of relationships between job ranks.

※ Analysis Methods

Here's how to use the Blockmodel algorithm in NetMiner:

1. In the top menu, go to Analyze > Position > Blockmodel (Conventional) > work interact, select Department in Select Vector, and click Run Process.

2. In the analysis results, navigate to the [T] Block Image Matrix tab, right-click, and select Save as new workfile to save the results as a new workfile.

3. In the top menu, go to Visualize > Spring > 2D, select [T] Block Sum Matrix, and click Run Layout.

4. Click the  icon at the top > Link Attribute Styling > Width > Select @WEIGHT > Click Apply.

※ Analysis Results

1. Finance Department

- It was identified as the department with the highest demand for collaboration.

- It has formed a particularly strong working relationship with the Marketing department.

- The Finance department has only one representative, so no internal relationships were formed.

2. Sales and Marketing Departments

- Both commonly exhibit a high number of internal working relationships.

- This indicates that relationships within the department are actively maintained.

These dynamics can be further confirmed with more precise figures in the matrix data. Double-click the [T] Block Sum Matrix table on the left side of the NetMiner screen. 

Click the Matrix tab at the bottom to view the number of links between departments.

3. Analysis of Organizational Members' Opinions

"So many diverse thoughts as there are members—could we periodically summarize just the key points in a simple way?"

The third analysis method involves using text mining to summarize employees' opinions and identify key insights. 

Using unstructured data, such as open-ended survey responses (e.g., "opinion") from the data examples we reviewed previously, NetMiner can perform a keyword network analysis. 

This can be used to summarize and analyze employees' thoughts effectively.

※ Analysis Methods

Here's how to conduct keyword network analysis in NetMiner:

1. Click File > Import Unstructured Text (available in NetMiner Premium).

2. In Input Excel/CSV File or Text Files, upload a file containing unstructured text.

3. Use Filter and Dictionary Settings to group synonyms or exclude unnecessary words.

4. Create a word cloud:

- Go to Tools > Plug-ins > Word Cloud (available in NetMiner Premium).

5. Generate a word network:

- Go to Tools > Plug-ins > Word Network, input the conditions, and create a network by linking words that appear in close proximity.

- Click Visualize > Spring > 2D > Run Layout to generate a word network map.

- Click Analyze > Centrality > Degree to identify the key keyword in the word network.

※ Analysis Results

1. It was found that employees mainly discussed topics such as ‘company,’ ‘communication,’ ‘leadership,’ and ‘career.’

2. It is evident that teamwork, communication, and leadership in the workplace are considered important.

3. Additionally, employees value whether their work life contributes to their personal career growth.

With NetMiner's text mining capabilities, it is possible to quantitatively identify key points. 

This can be used to gather opinions during organizational issues and summarize them to pinpoint the core insights. 

Additionally, by utilizing NetMiner's topic modeling features, you can categorize subtopics and quantitatively assess the proportion of each topic.

A course on text network analysis using NetMiner is now available on Udemy to help you gain a comprehensive understanding of the topic. 

In this course, you can explore the text network analysis methods introduced above in greater detail and practice them hands-on using NetMiner.

>> Course registration

We have introduced methods for conducting organizational network analysis using NetMiner. 

If you have any questions, please feel free to contact us via email.

Thank you!

※ The Text Analysis features is available in the NetMiner Premium edition.

※ Email : netminer@cyram.com