Graph Drawing with Morphing Partial Edges

Katsuya Akasaka; Kazuo Misue

REVIEW 3 major objections 4 minor 14 references

Graph Drawing with Morphing Partial Edges

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper proposes that animating the stubs of partial edge drawings—morphing them between partial and complete—can shorten the time readers need to confirm adjacency in a graph, and presents a scheduling algorithm plus a user study…

desk verdict A small, honest paper that adds time-dependence to partial edge drawing; the main speedup is credible on one task but the general conclusion outruns the single-speed experiment. read the letter →

arxiv 1908.01441 v5 pith:5DZDKVDC submitted 2019-08-05 cs.DS cs.CG

Kazuo Misue , Katsuya Akasaka This is my paper · ORCID

classification cs.DScs.CG

keywords graphdrawingpartialedgemorphingstub-edgeratiosymmetrichomogeneousMEDschedulingreadingtimeuserstudy

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proposes a time-varying version of partial edge drawing, called morphing edge drawing (MED), in which the stubs of each edge expand to a full line and shrink back on a schedule. Static partial edge drawings reduce visual clutter by omitting middle segments, but they force readers to infer the missing parts and therefore take longer to read. The paper formalizes MED as a morphing function on stub parameters, specifies scheduling requirements that prevent new crossings and shorten total morphing time, and gives an algorithm that assigns start times to edges. A user study with 12 participants finds that a symmetric homogeneous MED with stub ratio 1/4 significantly shortens the time to confirm whether two nodes are adjacent compared to the corresponding static PED, while accuracy differences did not reach significance. The paper concludes that morphing can act as a countermeasure to the extra reading time caused by partial edge drawing.

What carries the argument

The central object is the morphing function: for each edge $e$, a map $\mu_e$ from time to a partial drawing of $e$, written $\mu_e(t) = \gamma_e(\rho_e(t))$, where $\gamma_e$ is the stub-drawing function and $\rho_e$ is a ratio function selecting the two stub parameters $\alpha, \beta$ at time $t$. In a symmetric homogeneous MED with parameters $(\delta, \eta)$, every edge uses the same ratio function that moves the stub ratio from $\delta$ to $\eta$ and back; when $\eta = 1/2$ the stubs meet and the edge momentarily becomes complete. The scheduling mechanism is a morphing group: a connected component in the intersection graph of edges whose blank areas overlap. Start times are assigned in descending edge length; for each edge, forbidden start intervals are computed from the first and second times its stubs would pass crossing points of already scheduled intersecting edges, and the earliest allowed time is chosen. This packing of one-way travel times prevents dynamic stubs from crossing in the blank areas while shortening the total morphing cycle.

What would settle it

Run the same T1 adjacency task with morphing speed varied below and above 10 degrees per second (for example 5, 7, and 13 degrees per second) and with the 300 ms minimum removed; if the response-time advantage of 1/4-SHMED over 1/4-SHPED disappears or reverses at slower settings, the claim that morphing itself shortens reading time would be refuted. A second check is to record eye-tracking fixations on stub tips while two highlighted nodes are connected: if viewers do not track the moving tips, the proposed mechanism is not the cause of the measured benefit.

Watch

Extended reading notes

Core claim

The central claim is that animating partial edges—letting each stub grow to the full edge and shrink back—recovers some of the reading speed lost when a graph is drawn as a partial edge drawing, without sacrificing accuracy. More exactly, in an experiment with three displays (complete edges, 1/4-SHPED, and 1/4-SHMED) on a 50-node, 144-edge graph, the morphing version had a significantly shorter mean response time than the static partial version for the adjacency-confirmation task T1, placing it between the complete drawing and the static partial drawing. For the node-selection task T2 no significant response-time differences appeared, and no significant accuracy differences appeared in either task, so the hypothesis that MED is more accurate than CED was not supported. The paper interprets the result as evidence that morphing helps viewers connect stubs without requiring them to estimate the hidden middle.

Load-bearing premise

The experiment's chosen stub speed—10 degrees per second of visual angle, with a 300 ms minimum one-way travel time—is assumed to be fast enough to make morphing useful but slow enough for human smooth-pursuit eye movements to track the moving stub tips; participant comments that the stubs 'change too fast' suggest this assumption is near its limit.

Editorial extensions

If this is right

For the adjacency-check task T1, 1/4-SHMED achieved significantly shorter response times than 1/4-SHPED, with CED still fastest.
For the larger neighborhood-selection task T2, no significant response-time differences were found, so the morphing benefit did not generalize to all reading tasks in this experiment.
No significant accuracy differences appeared across CED, 1/4-SHPED, and 1/4-SHMED, so the paper's hypothesis H2 (MED more accurate than CED) is not supported.
The scheduling algorithm assigns start times so that dynamic stubs do not enter a blank area while another stub is there, avoiding new crossings, and packs the morphing intervals of non-intersecting edges to shorten total morphing time.
Taken together, the results support the paper's conclusion that MED can serve as a countermeasure to the extra time required to read a graph drawn with PED.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

An implication the authors leave implicit: because the speed choice came from smooth-pursuit limits, the time benefit is likely tied to the 10 degrees per second setting; at slower speeds the advantage may shrink, and several participants already reported the stubs moved too fast.
The study used one graph with 50 nodes and 144 edges and 12 participants, so the point estimate is thin; a replication across densities, stub ratios, and participant pools would show whether the adjacency-time advantage is stable.
A practical extension, not examined here, is selective morphing: animate only edges near the viewer's attention or only edges whose blank areas contain crossings, which could keep most of the benefit while reducing the visual clutter that participants complained about.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

First, know this: this is a small but honest contribution to graph drawing. The idea of making partial edge drawings time-dependent (MED) is a natural union of PED and animation, and the authors formalize it cleanly with ratio functions and morphing groups. They also ran a real user study with a Holm-corrected significance test, and the main positive result—1/4-SHMED cuts response time on the T1 adjacency task relative to 1/4-SHPED—looks credible. The scheduler heuristic is plausible for avoiding crossings, and the authors report participant complaints honestly.

The soft spots are real but not fatal. The stress-test hit the main one: everything hangs on a single morphing speed (10 degrees/s with a 300 ms floor), chosen from smooth-pursuit data, and participants already said 'the stubs change too fast.' Without a speed sweep, H1 is one point on the curve, not a general property of MED. The conclusion in Section 7 overreaches when it says MED 'can function as a countermeasure' against PED reading time, given only one of two tasks was significant and H2 (accuracy) was not supported. Also, the morphing-group scheduler has no proof of correctness; for a visualization paper that's acceptable if the heuristic is shown to work, but there's no analysis of how often it finds a feasible schedule. Minor points: n=12, no code/data released, and the response-time distributions are shown only as boxplots without individual data.

Who is this for? People who work on partial edge drawing, readability of node-link diagrams, or animated edge encodings. It won't change the field, but it's a legitimate variation with a testable hypothesis and an honest experiment. It deserves peer review—not desk rejection. I'd ask for a speed sweep or a heavily softened conclusion, plus release of the materials, but the core idea is clear and worth engaging.

Referee Report

3 major / 4 minor

Summary. The paper introduces morphing edge drawing (MED), a time-varying variant of partial edge drawing (PED), with a formalization of PED/MED, two scheduling requirements (R1: avoid new crossings in blank areas; R2: shorten total morphing time), a greedy scheduling algorithm (Algorithm 1), and a user study comparing CED, 1/4-SHPED, and 1/4-SHMED on two tasks. The main empirical claim is that 1/4-SHMED significantly shortens response time for adjacency verification (T1) compared with 1/4-SHPED (p=0.0011 after Holm correction), although no significant advantage is found for neighbor enumeration (T2) or accuracy. The paper concludes that MED can function as a countermeasure against the reading time incurred by PED.

Significance. The formalization in Sections 2 and 4 is a clean extension of Bruckdorfer et al.'s PED model and is a useful conceptual contribution. The scheduling requirements R1/R2 are sensible, and the experiment uses an appropriate nonparametric framework (Friedman test with Holm correction). If the T1 result is robust across morphing speeds, the paper would demonstrate a practical way to recover reading speed lost by PED. The main limitations—unproven schedule correctness, a single morphing speed, and a small n with only one significant task—currently prevent the broader conclusion from being fully supported.

major comments (3)

[Section 5.5, Algorithm 1] The algorithm has no correctness proof. The manuscript asserts that the schedule satisfies R1 and R2, but it does not prove that respecting the pairwise prohibited intervals from Expressions (3)-(4) prevents stub crossings throughout an entire morphing cycle, nor does it account for periodic repetition of morphing. It also does not define a measure for R2 or prove any optimality or approximation property. Because the scheduling algorithm is part of the contribution and is used in the experiment, this gap is load-bearing.
[Sections 6.4 and 6.9] The experiment fixes morphing speed at 10 deg/s with a 300 ms minimum one-way travel time. Participant feedback in Section 6.9 explicitly states "The stubs change too fast. The time for stubs to connect is too short," and the authors acknowledge that "the morphing speed is too fast." Thus H1 is supported at a single speed that is at the usability boundary, and the conclusion in Section 7 extrapolates to MED in general. A speed sweep (for example, 5 and 15 deg/s or different minimum travel times) is needed to establish that the T1 advantage is a property of MED rather than an artifact of this speed.
[Sections 6.5 and 7, Tables 1-3] With 12 participants, only the T1 response-time comparison shows a significant effect; T2 response time and accuracy show no significant differences. The concluding sentence in Section 7, "MED can function as a countermeasure against the time to read a graph by PED," is broader than the evidence, which supports only a speed advantage for the adjacency-verification task at one morphing speed. The conclusion should be restricted accordingly or supported by additional tasks and speeds.

minor comments (4)

[Section 2] 'asymmetric PED (SPED)' should be 'symmetric PED (SPED)'; the condition alpha_e = 1 - beta_e gives equal-length stubs, and the same abbreviation is used in Section 4.1.
[Section 6.9] 'morphing of two stabs works simultaneously' should read 'two stubs'.
[Tables 1-3] State explicitly whether the reported p-values are raw or Holm-adjusted, and provide the Friedman test statistics.
[Section 5] Define 'blank area' formally at the point where R1 is introduced, since the algorithm and morphing groups rely on it.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the MED formalism and scheduling are derived from graph geometry, and the user study is an independent external test.

full rationale

The paper contains no derivation step in which a predicted quantity is equivalent by construction to an input parameter. The morphing ratio function (Exp. 2) is defined from the chosen delta, eta, edge length l, constant speed s, and times t0/t1/t2; these parameters are set before the experiment, and no response-time or accuracy data are used to refit them. The scheduling algorithm (Algorithm 1 and Exps. 3-5) computes start times from geometric crossing intervals only. The central empirical claim (H1, Table 1) is assessed by a controlled comparison of CED, 1/4-SHPED, and 1/4-SHMED on independent participants, so the outcome is externally falsifiable rather than forced by the formalization. The morphing-speed choice and participant feedback about speed are external-validity concerns, not circularity. There are no load-bearing self-citations or imported uniqueness theorems. Score 0.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new physical entities. Its main unstated costs are hand-set experimental parameters (morphing speed, minimum travel time, stub ratio) and an unproved correctness assumption for the greedy scheduler. All other assumptions are standard domain assumptions about fixed straight-line layouts and constant speed.

free parameters (3)

stub morphing speed = 10 degrees/s
Hand-set in Section 6.4 using Robinson's smooth-pursuit data; it directly determines all morphing timings and the reading-time measurement.
minimum one-way travel time = 300 ms
Hand-set floor in Section 6.4 to keep stubs trackable; constrains the schedule and may mask faster or slower effects.
stub-edge ratio delta = 1/4
Inherited from prior 1/4-SHPED studies; a design choice, not fitted, but it defines the experimental conditions.

assumptions (5)

domain assumption The straight-line layout Gamma(G) is fixed in advance and stub crossings are allowed.
Section 2 states this assumption; all scheduling operates on a given drawing and does not repair layout-level crossings.
domain assumption All stubs morph at the same constant speed.
Section 4.1, Eq. (2) fixes uniform speed; used by the greedy scheduler in Algorithm 1.
standard math Two edges that do not cross cannot create new stub crossings during morphing.
Section 5.1 uses this to partition edges into morphing groups; it follows from stubs being subsegments of straight-line edges.
ad hoc to paper Algorithm 1's greedy earliest-start-time packing yields a schedule satisfying R1 and R2.
Section 5.5 presents Algorithm 1 without a correctness proof; this unproved algorithmic premise underpins the claim that morphing can be scheduled without new crossings.
domain assumption A morphing speed of 10 degrees/s is trackable by human smooth pursuit based on Robinson [12].
Section 6.4 sets the experimental speed from Robinson's eye-movement data; no validation in this paper's setup.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Graph Drawing with Morphing Partial Edges." pith.science (2026). https://pith.science/paper/5DZDKVDC

@misc{pith2026190801441,
  author       = {Pith},
  title        = {Pith review of: Graph Drawing with Morphing Partial Edges},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5DZDKVDC}},
  note         = {Machine review of arXiv:1908.01441}
}

read the original abstract

A partial edge drawing (PED) of a graph is a variation of a node-link diagram. PED draws a link, which is a partial visual representation of an edge, and reduces visual clutter of the node-link diagram. However, more time is required to read a PED to infer undrawn parts. The authors propose a morphing edge drawing (MED), which is a PED that changes with time. In MED, links morph between partial and complete drawings; thus, a reduced load for estimation of undrawn parts in a PED is expected. Herein, a formalization of MED is shown based on a formalization of PED. Then, requirements for the scheduling of morphing are specified. The requirements inhibit morphing from crossing and shorten the overall time for morphing the edges. Moreover, an algorithm for a scheduling method implemented by the authors is illustrated and the effectiveness of PED from a reading time viewpoint is shown through an experimental evaluation.

Figures

Figures reproduced from arXiv: 1908.01441 by the authors.

**Figure 1.** shows the graph of the function. ρe(t) =    δe for t ≤ t0 or t2 < t δe + (t − t0)s/l for t0 < t ≤ t1 ηe − (t − t1)s/l for t1 < t ≤ t2, (2) where t1 is the time when the stub-edge ratio becomes ηe, and t2 is the time when the stub-edge ratio returns to δe. Using one-way travel time d1 = (ηe − δe)l/s, t1 = t0 + d1 and t2 = t0 + 2d1. 1 1 − ηe 1 − δe ηe δe 0 ρe(t) t0 t1 t2 t [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗

**Figure 2.** Dependency between edges. The edges e1 and e2 do not intersect, but both intersect e3, so they cannot be morphed independently. Dotted lines represent the omitted parts (blank areas). A set of edges where the timing of morphing may affect each other is a morphing group. Different morphing groups can be scheduled independently. To determine the morphing groups, another graph is generated from the graph to be drawn. T… view at source ↗

**Figure 3.** Packing morphing intervals. 5.4 Parallel Morphing Even if edges belong to the same morphing group, morphing of two non-intersecting edges may be performed simultaneously. By appropriately morphing parallelly, it is possible to shorten the total morphing time of all edges. For example, as edges e1 and e2 in [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: Examples of visual representations used in T1 (a) CED (b) 1/4-SHPED (c) 1/4-SHMED [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]

**Figure 5.** Figure 5: Examples of visual representations used in T2 6.3 Graphs used for the experiment We used the Barab´asi-Albert model [1] as a guideline to create a graph with 50 nodes and 144 edges. We used the Fruchterman-Reingold algorithm [10] to determine the layout of the graph. 6…

**Figure 6.** Figure 6: Distribution of response time [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]

**Figure 7.** Figure 7: Correct answer rate • It is messy and difficult to see. My eyes are strained. • The stubs change too fast. The time for stubs to connect is too short. Positive opinions indicate that morphing contributes to reading graphs. In contrast, from the negative opinions, it ap…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

14 extracted references · 8 canonical work pages

[1]

Science 286(5439), 509–512 (1999)

Barab´ asi, A.L., Albert, R.: Emergence of scaling in random networks. Science 286(5439), 509–512 (1999). https://doi.org/10.1126/science.286.5439.509 Graph Drawing with Morphing Partial Edges 13

doi:10.1126/science.286.5439.509 1999
[2]

IEEE Trans- actions on Visualization and Computer Graphics 1(1), 16–28 (March 1995)

Becker, R.A., Eick, S.G., Wilks, A.R.: Visualizing network data. IEEE Trans- actions on Visualization and Computer Graphics 1(1), 16–28 (March 1995). https://doi.org/10.1109/2945.468391

arXiv 1995
[3]

In: 2016 7th International Con- ference on Information, Intelligence, Systems Applications (IISA)

Binucci, C., Liotta, G., Montecchiani, F., Tappini, A.: Partial edge drawing: Homo- geneity is more important than crossings and ink. In: 2016 7th International Con- ference on Information, Intelligence, Systems Applications (IISA). pp. 1–6 (July 2016). https://doi.org/10.1109/IISA.2016.7785427

arXiv 2016
[4]

IEEE Trans- actions on Visualization and Computer Graphics 15(6), 969–976 (Nov 2009)

Blaas, J., Botha, C., Grundy, E., Jones, M., Laramee, R., Post, F.: Smooth graphs for visual exploration of higher-order state transitions. IEEE Trans- actions on Visualization and Computer Graphics 15(6), 969–976 (Nov 2009). https://doi.org/10.1109/TVCG.2009.181

work page doi:10.1109/tvcg.2009.181 2009
[5]

(eds.) Fun with Algorithms

Bruckdorfer, T., Kaufmann, M.: Mad at edge crossings? break the edges! In: Kranakis, E., Krizanc, D., Luccio, F. (eds.) Fun with Algorithms. pp. 40–50. Springer Berlin Heidelberg, Berlin, Heidelberg (2012)

work page 2012
[6]

In: Di Giacomo, E., Lubiw, A

Bruckdorfer, T., Kaufmann, M., Leibßle, S.: PED user study. In: Di Giacomo, E., Lubiw, A. (eds.) Graph Drawing and Network Visualization. pp. 551–553. Springer International Publishing, Cham (2015)

work page 2015
[7]

In: 2017 21st International Conference Information Visualisation (iV)

Burch, M.: A user study on judging the target node in partial link drawings. In: 2017 21st International Conference Information Visualisation (iV). pp. 199–204 (July 2017). https://doi.org/10.1109/iV.2017.43

work page doi:10.1109/iv.2017.43 2017
[8]

In: van Kreveld, M., Speckmann, B

Burch, M., Vehlow, C., Konevtsova, N., Weiskopf, D.: Evaluating partially drawn links for directed graph edges. In: van Kreveld, M., Speckmann, B. (eds.) Graph Drawing. pp. 226–237. Springer Berlin Heidelberg, Berlin, Heidelberg (2012)

work page 2012

Show all 14 references

[9]

In: 2009 IEEE Symposium on Visual Analytics Science and Technology

Collins, C., Vi´ egas, F.B., Wattenberg, M.: Parallel tag clouds to ex- plore and analyze faceted text corpora. In: 2009 IEEE Symposium on Visual Analytics Science and Technology. pp. 91–98 (Oct 2009). https://doi.org/10.1109/VAST.2009.5333443

2009
[10]

Software: Practice and Experience 21(11), 1129–1164 (1991)

Fruchterman, T.M.J., Reingold, E.M.: Graph drawing by force-directed placement. Software: Practice and Experience 21(11), 1129–1164 (1991). https://doi.org/10.1002/spe.4380211102

1991 doi
[11]

In: 2011 IEEE Paciﬁc Visualization Symposium

Holten, D., Isenberg, P., van Wijk, J.J., Fekete, J.: An extended evaluation of the readability of tapered, animated, and textured directed-edge representations in node-link graphs. In: 2011 IEEE Paciﬁc Visualization Symposium. pp. 195–202 (March 2011). https://doi.org/10.1109...

2011
[12]

The Journal of Physiology 180(3), 569–591 (1965)

Robinson, D.A.: The mechanics of human smooth pursuit eye movement. The Journal of Physiology 180(3), 569–591 (1965). https://doi.org/10.1113/jphysiol.1965.sp007718

1965 doi
[13]

In: Proceedings of the 2018 CHI Conference on Human Factors in Com- puting Systems

Romat, H., Appert, C., Bach, B., Henry-Riche, N., Pietriga, E.: Animated edge textures in node-link diagrams: A design space and initial evaluation. In: Proceedings of the 2018 CHI Conference on Human Factors in Com- puting Systems. pp. 187:1–187:13. CHI ’18, ACM, New York, NY...

2018
[14]

In: Proceedings of 6th International Conference on Information Visualization Theory and Applications (IVAPP)

Schmauder, H., Burch, M., Weiskopf, D.: Visualizing dynamic weighted digraphs with partial links. In: Proceedings of 6th International Conference on Information Visualization Theory and Applications (IVAPP). pp. 123–130 (2015)

2015

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Science 286(5439), 509–512 (1999)

Barab´ asi, A.L., Albert, R.: Emergence of scaling in random networks. Science 286(5439), 509–512 (1999). https://doi.org/10.1126/science.286.5439.509 Graph Drawing with Morphing Partial Edges 13

doi:10.1126/science.286.5439.509 1999

[2] [2]

IEEE Trans- actions on Visualization and Computer Graphics 1(1), 16–28 (March 1995)

Becker, R.A., Eick, S.G., Wilks, A.R.: Visualizing network data. IEEE Trans- actions on Visualization and Computer Graphics 1(1), 16–28 (March 1995). https://doi.org/10.1109/2945.468391

arXiv 1995

[3] [3]

In: 2016 7th International Con- ference on Information, Intelligence, Systems Applications (IISA)

Binucci, C., Liotta, G., Montecchiani, F., Tappini, A.: Partial edge drawing: Homo- geneity is more important than crossings and ink. In: 2016 7th International Con- ference on Information, Intelligence, Systems Applications (IISA). pp. 1–6 (July 2016). https://doi.org/10.1109/IISA.2016.7785427

arXiv 2016

[4] [4]

IEEE Trans- actions on Visualization and Computer Graphics 15(6), 969–976 (Nov 2009)

Blaas, J., Botha, C., Grundy, E., Jones, M., Laramee, R., Post, F.: Smooth graphs for visual exploration of higher-order state transitions. IEEE Trans- actions on Visualization and Computer Graphics 15(6), 969–976 (Nov 2009). https://doi.org/10.1109/TVCG.2009.181

work page doi:10.1109/tvcg.2009.181 2009

[5] [5]

(eds.) Fun with Algorithms

Bruckdorfer, T., Kaufmann, M.: Mad at edge crossings? break the edges! In: Kranakis, E., Krizanc, D., Luccio, F. (eds.) Fun with Algorithms. pp. 40–50. Springer Berlin Heidelberg, Berlin, Heidelberg (2012)

work page 2012

[6] [6]

In: Di Giacomo, E., Lubiw, A

Bruckdorfer, T., Kaufmann, M., Leibßle, S.: PED user study. In: Di Giacomo, E., Lubiw, A. (eds.) Graph Drawing and Network Visualization. pp. 551–553. Springer International Publishing, Cham (2015)

work page 2015

[7] [7]

In: 2017 21st International Conference Information Visualisation (iV)

Burch, M.: A user study on judging the target node in partial link drawings. In: 2017 21st International Conference Information Visualisation (iV). pp. 199–204 (July 2017). https://doi.org/10.1109/iV.2017.43

work page doi:10.1109/iv.2017.43 2017

[8] [8]

In: van Kreveld, M., Speckmann, B

Burch, M., Vehlow, C., Konevtsova, N., Weiskopf, D.: Evaluating partially drawn links for directed graph edges. In: van Kreveld, M., Speckmann, B. (eds.) Graph Drawing. pp. 226–237. Springer Berlin Heidelberg, Berlin, Heidelberg (2012)

work page 2012