Sequence Alignment of Behavioural Paths

Description

Measures similarity between users' action sequences by edit cost, then clusters paths to reveal typical routes and deviant trajectories.

Listed in the treatise's borrowing matrix, sequence alignment comes from bioinformatics (aligning DNA sequences) and was adapted to social science by Abbott as optimal matching. The distance between two sequences is the minimum cost of insertions, deletions and substitutions to turn one into the other. Pairwise distances are then clustered to produce typologies of trajectories. In UX it compares whole navigation paths rather than single transitions: which routes do users take to complete onboarding, and how do failing paths differ from successful ones? Cost settings must be justified, and results visualised with sequence index plots.

Attributes

Loss type
Unmigrated
Stage
Post-launch
Effort
Days

Steps

  1. Encode sessions or journeys as sequences of states or actions.
  2. Define substitution and indel costs (data-driven or theory-based).
  3. Compute pairwise distances with optimal matching or alternatives (for example LCS).
  4. Cluster sequences and select a typology using quality indices.
  5. Visualise with sequence index and state distribution plots.
  6. Relate typology to outcomes and investigate deviant paths qualitatively.

Sources

  • Lost Methods in UX — A Complete Treatise (Gopinath)
  • Abbott A & Tsay A (2000) Sequence analysis and optimal matching methods in sociology. Sociological Methods & Research 29(1)
  • Gabadinho A et al. (2011) Analyzing and visualizing state sequences in R with TraMineR. J Stat Software
  • Abbott & Tsay, Sequence analysis and optimal matching methods in sociology (2000), Sociological Methods & Research
  • Gabadinho et al., Analyzing and visualizing state sequences in R with TraMineR (2011), Journal of Statistical Software
  • Cornwell, Social Sequence Analysis (2015)

Also known as

  • Optimal matching analysis
  • Sequence analysis
  • Clickstream alignment
  • Social sequence analysis
  • TraMineR analysis
  • Sequence Analysis (Optimal Matching)

Categories

  • Analytics & Behaviour
  • Computational & Machine Learning