Usage Data Analytics

Software analytics

• enables software practitioners

  • to explore and analyze data

    • in order to obtain insightful and actionable information

      • for data driven tasks

• decision based on facts

• increased transparency and traceability

• less subjective decision making

• automated way for decision making

• decreased risk of “overlooking”

Use of data —> Create insight and actionable information to support decisions

Data + Data analytics processes —> Insight

• Descriptive Analytics “What has happened?”

• Diagnostic Analtytics “Why has happened?”

• Predictive Analytics “What may happen?”

• Prescriptive Analytics “How to prevent things from happening?”

• Use of data to identify trends and relationships

• Ad-hoc reporting and dashboards

• Focuses on questions

  • what has happened?, how does it compare to our plan?, what is happening now?

• e.g. Google analytics

• Use of data to determine the causes of trends and correlations between variables

• Deeper analysis on ause of phenomena

• Focuses on question

  • Why has it happened?

• Challenges

  • Laborious work

• Use of data, statistical algorithms and machine learning techniques

  • to identify likelihood of future outcomes

• Focuses on question

  • What may happen?

• Data needed:

  • Contextual and Interaction Data

• Techniques needed:

  • Advanced statistics, Modeling, Data Mining, Machine Learning

• Use of data to determine an optimal course of action

• Focuses on questions

  • How prevent things from happening?

  • How make sure specific things happens as we want?

• Data needed: Contextual, Interaction & Predicted Data

• Techniques needed: Advanced statistics, Modeling, Data mining, Machine Learning

Stakeholder

• Developers

• Managers

• Tester

• Dev. Lead

• Test Lead

• Researchers

• Users

• Questions

  
  

  Data Analytic

  
  

• Answers

Study

• 17 SaaS product managers

• Interviewees rated importance level of analytics for decision

  • regarding portfolio management, roadmapping and release planning

• Software Engineering Practitioners —> Historical Repositories (Mailing Lists, Bug Tracking)

• Users —> Runtime Repositories (Logs, Comments)

• Use Visualization to make it easy

• Be precise

• Measure many artifacts

• Identify important & unusual items

Data about users interaction while using system

• Context data

  • info that provides a broader understanding of event, person or item

• Interaction data

  • record of actions taken by user with application

• Clustering

• Pattern recognition

• Data Mining

• Information cisualization

• Classification

• Goal

  • process natural language texts and to make their information analyzable to computer systems

• Techniques

  • Indexing

  • Pattern Analysis

  • Collocations

are used to present analytical results visually

• many patterns only recognized by stakeholders

• meaning of data understood easier when presented as chart/graph

2 types of data cisualization

• Static visualization

• Interactive visualization

  • stakeholder is able to interact with visualization