Massimo Franceschet

PageRank: standing on the shoulders of giants

The roots of Googles PageRank can be traced back to several early, and equally remarkable, ranking techniques.

A comparison of bibliometric indicators for computer science scholars and journals on Web of Science and Google Scholar

Given the current availability of different bibliometric indicators and of production and citation data sources, the following two questions immediately arise: do the indicators’ scores differ when computed on different data sources? More importantly, do the indicator-based rankings significantly change when computed on different data sources? We provide a case study for computer science scholars and journals evaluated on Web of Science and Google Scholar databases. The study concludes that Google scholar computes significantly higher indicators’ scores than Web of Science. Nevertheless, citation-based rankings of both scholars and journals do not significantly change when compiled on the two data sources, while rankings based on the h index show a moderate degree of variation.

The effect of scholar collaboration on impact and quality of academic papers

We study how scholar collaboration varies across disciplines in science, social science, arts and humanities and the effects of author collaboration on impact and quality of co-authored papers. Impact is measured with the aid of citations collected by papers, while quality is determined by the judgements expressed by peer reviewers. To this end, we take advantage of the dataset provided by the first-ever national research assessment exercise of Italian universities, which involved 20 disciplinary areas, 102 research structures, 18,500 research products, and 6661 peer reviewers. Collaboration intensity neatly varies across disciplines: it is inescapable is most sciences and negligible in most humanities. We measured a general positive association between cardinality of the author set of a paper and citation count as well as peer quality of the contribution. The correlation is stronger when the affiliations of authors are heterogeneous. There exist, however, notable and interesting counter-examples.

XPathMark: an XPath benchmark for the XMark generated data

We propose XPathMark, an XPath benchmark on top of the XMark generated data. It consists of a set of queries which covers the main aspects of the language XPath 1.0. These queries have been designed for XML documents generated under XMark, a popular benchmark for XML data management. We suggest a methodology to evaluate the XPathMark on a given XML engine and, by way of example, we evaluate two popular XML engines using the proposed benchmark.

Representing and Reasoning about Temporal Granularities

In this paper, we propose a new logical approach to represent and to reason about different time granularities. We identify a time granularity as an infinite sequence of time points properly labelled with proposition symbols marking the starting and ending points of the corresponding granules, and we symbolically model sets of granularities by means of linear time logic formulas. Some real-world granularities are provided, from a clinical domain and from the Gregorian Calendar, to motivate and exemplify our approach. Different formulas are introduced, which represent relations between different granularities. The proposed framework permits one to algorithmically solve the consistency, the equivalence, and the classification problems in a uniform way, by reducing them to the validity problem for the considered linear time logic.

The first Italian research assessment exercise: A bibliometric perspective

In December 2003, seventeen years after the first UK research assessment exercise, Italy started up its first-ever national research evaluation, with the aim to evaluate, using the peer review method, the excellence of the national research production. The evaluation involved 20 disciplinary areas, 102 research structures, 18,500 research products and 6661 peer reviewers (1465 from abroad); it had a direct cost of 3.55 millions Euros and a time length spanning over 18 months. The introduction of ratings based on ex post quality of output and not on ex ante respect for parameters and compliance is an important leap forward of the national research evaluation system toward meritocracy. From the bibliometric perspective, the national assessment offered the unprecedented opportunity to perform a large-scale comparison of peer review and bibliometric indicators for an important share of the Italian research production. The present investigation takes full advantage of this opportunity to test whether peer review judgements and (article and journal) bibliometric indicators are independent variables and, in the negative case, to measure the sign and strength of the association. Outcomes allow us to advocate the use of bibliometric evaluation, suitably integrated with expert review, for the forthcoming national assessment exercises, with the goal of shifting from the assessment of research excellence to the evaluation of average research performance without significant increase of expenses.

Model Checking Hybrid Logics (With An Application to Semistructured Data)

We investigate the complexity of the model checking problem for hybrid logics. We provide model checker algorithms for various hybrid fragments and we prove PSPACE-completeness for hybrid fragments including binders. We complement and motivate our complexity results with an application of model checking in hybrid logic to the problems of query and constraint evaluation for semistructured data.

Collaboration in computer science: A network science approach

We represent collaboration of authors in computer science papers in terms of both affiliation and collaboration networks and observe how these networks evolved over time since 1960. We investigate the temporal evolution of bibliometric properties, like size of the discipline, productivity of scholars, and collaboration level in papers, as well as of large-scale network properties, like reachability and average separation distance among scientists, distribution of the number of scholar collaborators, network clustering and network assortativity by number of collaborators.

On the complexity of hybrid logics with binders

Hybrid logic refers to a group of logics lying between modal and first-order logic in which one can refer to individual states of the Kripke structure. In particular, the hybrid logic HL(@, ↓ ) is an appealing extension of modal logic that allows one to refer to a state by means of the given names and to dynamically create new names for a state. Unfortunately, as for the richer first-order logic, satisfiability for the hybrid logic HL(@, ↓ ) is undecidable and model checking for HL(@, ↓ ) is PSpace-complete. We carefully analyze these results and we isolate large fragments of HL(@, ↓ ) for which satisfiability is decidable and model checking is below PSpace.

Journal influence factors

We performed a thorough comparison of four main indicators of journal influence, namely 2-year impact factor, 5-year impact factor, eigenfactor and article influence. These indicators have been recently added by Thomson Reuters to the Journal Citation Reports, in both science and social science editions, and are thus available for study and comparison over a sample of significative size. We find that the distribution associated with the eigenfactor largely differs from the distribution of the other surveyed measures in terms of deviation from the mean, concentration, entropy, and skewness. Moreover, it is the one that best fits to the lognormal theoretical model. Surprisingly, the eigenfactor is also the most variable indicator when computed across different fields of science and social science, while article influence is the most stable in this respect, and hence the most suitable metric to be used interdisciplinarily. Finally, the journal rankings provided by impact factors and article influence are relatively similar and diverge from the one produced by eigenfactor, which is closer to that given by the total number of received citations.