About Me

I completed my undergraduate degree at UCL in 2012 in Computer Science. I worked with others on the Newton Spectrum corpus browser, and then extended the system to incorporate a topic modelling based cluster visualisation. In my final year I worked on an automated method for correcting topic model coherence through external bias. I am currently a PhD candidate in Computer Science at UCL Crest centre, in the area of App Store Analysis.

Part of UCLappA group.
Contact me at w.martin (at) ucl.ac.uk

Research

<

A Survey of App Store Analysis for Software Engineering

--Accepted for publication in TSE-- Download | Bibtex
Abstract: App Store Analysis studies information about applications obtained from app stores. App stores provide a wealth of information derived from users that would not exist had the applications been distributed via previous software deployment methods. App Store Analysis combines this non-technical information with technical information to learn trends and behaviours within these forms of software repositories. Findings from App Store Analysis have a direct and actionable impact on the software teams that develop software for app stores, and have led to techniques for requirements engineering, release planning, software design, security and testing. This survey describes and compares the areas of research that have been explored thus far, drawing out common aspects, trends and directions future research should take to address open problems and challenges.

Causal Impact Analysis for App Releases in Google Play

--Accepted for publication at FSE 2016-- Download | Bibtex
Abstract: App developers would like to understand the impact of their own and their competitors’ software releases. To address this we introduce Causal Impact Release Analysis for app stores, and our tool, CIRA, that implements this analysis. We mined 38,858 popular Google Play apps, over a period of 12 months. For these apps, we identified 26,339 releases for which there was adequate prior and posterior time series data to facilitate causal impact analysis. We found that 33% of these releases caused a statistically significant change in user ratings. We use our approach to reveal important characteristics that distinguish causal significance in Google Play. To explore the actionability of causal impact analysis, we elicited the opinions of app developers: 56 companies responded, 78% concurred with the causal assessment, of which 33% claimed that their company would consider changing its app release strategy as a result of our findings.

Causal Impact for App Store Analysis

--Accepted for publication at the ACM Student Research competition at ICSE 2016-- Download | Poster | Research note | Bibtex
Abstract: App developers naturally want to know which of their releases are successful and which are unsuccessful. Such information can help with release planning and requirements prioritisation and elicitation. To address this problem, I performed causal analysis on 52 weeks of popular app releases from Google Play and Windows Phone Store. The results reveal properties of successful releases in multiple app stores, and showcase causal analysis as a useful technique for developers seeking to know more about their software releases.

The App Sampling Problem for App Store Mining

--Accepted for publication at MSR 2015-- Download | Bibtex
Abstract: Many papers on App Store Mining are susceptible to the App Sampling Problem, which exists when only a subset of apps are studied, resulting in potential sampling bias. We introduce the App Sampling Problem, and study its effects on sets of user review data. We investigate the effects of sampling bias, and techniques for its amelioration in App Store Mining and Analysis, where sampling bias is often unavoidable. We mine 106,891 requests from 2,729,103 user reviews and investigate the properties of apps and reviews from 3 different partitions: the sets with fully complete review data, partially complete review data, and no review data at all. We find that app metrics such as price, rating, and download rank are significantly different between the three completeness levels. We show that correlation analysis can find trends in the data that prevail across the partitions, offering one possible approach to App Store Analysis in the presence of sampling bias.

App Store Analysis: Mining App Stores for Relationships between Customer, Business and Technical Characteristics

Download | Bibtex
Abstract: This paper argues that App Store Analysis can be used to understand the rich interplay between app customers and their developers. We use data mining to extract price and popularity information and natural language processing and data mining to elicit each app’s claimed features from the Blackberry App Store, revealing strong correlations between customer rating and popularity (rank of app downloads). We found evidence for a mild correlation between price and the number of features claimed for an app and also found that higher priced features tended to be lower rated by their users. We also found that free apps have significantly (p-value < 0.001) higher rating than non-free apps, with a moderately high effect size (Â12 = 0.68). We also provide initial evidence that extracted claimed features are meaningful to developers (precision = 0.71, recall = 0.77). All data from our experiments and analysis are made available on-line to support further analysis.

Automatic Correction of Topic Coherence

Download | Bibtex
Abstract: A set of texts is often a poor representation of the language it is written in, and resultantly topics can seem nonsensical to domain experts. This can be for several reasons: misspellings or ‘accidental words’ can be given statistical significance in the case that too many topics are learned; words can appear related or unrelated in the text, even though the opposite is true in the language; too few topics or too many topics are used. In this position paper we present a novel approach by applying biases derived from external sources during the training process, in order to improve the coherence of topics. This has the effect of improving topic coherence [Newman et al., 2009, 2010], ironing out many of the issues that a sub-optimal number of topics can cause, and imbuing resultant models with real-world word-relationships.

A Performance Analysis of Distributed Indexing using Terrier

Download | Bibtex
Abstract: High performance indexing is critical to fast and efficient data retrieval, and underlies mainstream systems such as search engines. A limiting factor for indexing performance is data set size: while this may not be important for small desktop or smartphone applications, the scale of the dataset is problematic when indexing large corpora such as the web. Distributing both the dataset and the computation can solve this issue, and consequently a number of distributed indexers have been developed, but their effectiveness remains to be seen. To shed some light on this issue, we present a detailed performance analysis of one of the leading solutions. Terrier is a scalable piece of software supporting indexing and retrieval both on a single nodes and a distributed cluster. We demonstrate that performance is highly correlated to the specific setup of a cluster and aim to provide useful guidelines for data scientists willing to get the most out of their hardware.

Newton Spectrum

Interactive corpus browsing software written in Java. The corpus browser of the 21st century, built as a viewer, guide and exploration tool for use with large corpora such as Newton's corpus. Built as part of an undergraduate project with colleages from UCL in collaboration with the Newton Project (University of Sussex). The browser uses latent Dirichlet allocation in order to automatically analyse and cluster documents.

Research Interests

  • App Store repository mining and analysis
  • Longitudinal studies inc. causal inference
  • Unstructured textual analysis through latent topics
  • Clustering and classification of free-text documents

Publications

  • William Martin, Federica Sarro, Yue Jia, Yuanyuan Zhang and Mark Harman, "A Survey of App Store Analysis for Software Engineering", Transactions of Software Engineering, to appear.
  • William Martin, Federica Sarro and Mark Harman, "Causal Impact Analysis for App Releases in Google Play", FSE 2016, Seattle, Washington, USA.
    Paper | Bibtex | Website
  • William Martin, "Causal Impact for App Store Analysis", ACM Student Research Competition (SRC) 2016, Austin, Texas, USA.
    Poster | Paper | Bibtex | Data
  • William Martin, Federica Sarro, Yue Jia, Yuanyuan Zhang and Mark Harman, "A Survey of App Store Analysis for Software Engineering", Technical report.
    Paper | Bibtex
  • William Martin, Federica Sarro and Mark Harman, "Causal Impact Analysis Applied to App Releases in Google Play and Windows Phone Store", Technical report.
    Paper | Bibtex
  • William Martin, Mark Harman, Yue Jia, Federica Sarro and Yuanyuan Zhang, "The App Sampling Problem for App Store Mining", MSR 2015, Florence, Italy.
    Paper | Bibtex | Data
  • Anthony Finkelstein, Mark Harman, Yue Jia, William Martin, Federica Sarro and Yuanyuan Zhang, "App Store Analysis: Mining App Stores for Relationships between Customer, Business and Technical Characteristics", Technical report.
    Paper | Bibtex
  • William Martin and John Shawe-Taylor, "Automatic Correction of Topic Coherence", Technical report.
    Paper | Bibtex
  • Amaury Couste, Jakub Kozłowski and William Martin, "A Performance Analysis of Distributed Indexing using Terrier", Technical report.
    Paper | Bibtex

Contact Me